Do You Have a Dragon?
Marco:
If you want to like really screw anything up big time, like really, if you want to have just really make some of the biggest public mistakes of your life, now is the time to do it because at least we're not the open AI board.
Casey:
You're jumping ahead, my friend.
Casey:
Oh, jumping ahead.
Marco:
What a mess.
Marco:
Oh, my God.
Marco:
Like this.
Marco:
Wow.
Marco:
I mean, you know, you talk about people having having bad weeks like sometimes.
Marco:
Oh, yeah.
Marco:
They're having a bad week.
John:
boy are they having a bad week at least it's at least it's relatively low stakes i feel like it's it's it's fun because it's only like whatever 80 billion dollars on the line but it's you know in the end it's people with computers and stuff and things and it's not like life or death and most of the people involved you feel okay laughing at them a little bit because they're such big distant figures they don't seem like real people to you but rest assured they are and that's probably why they're in this mess
Casey:
Final warning.
Casey:
This is the moment where you, a dear, beloved listener who has not yet put in your order, you sit or stand there and you think, I've got this.
Casey:
No worries.
Casey:
I'm going to take care of that as soon as I get to the office or I'll take care of it as soon as I get home.
Casey:
And then you, dear, beloved listener.
Casey:
You forget.
Casey:
And then what happens?
Casey:
Then the rules state you have to go tweeting, even though I won't see it, or tooting at me saying, I'm the one.
Casey:
I'm that person this time.
Casey:
Don't be that person.
Casey:
Don't do that to yourself.
Casey:
Go to atp.fm.
Casey:
This is your final warning.
Casey:
This is all we got.
Casey:
This is it.
Casey:
The sale ends this weekend, Sunday, November 26th.
Casey:
That's it.
Casey:
That's all the time you get.
Casey:
So go now.
Casey:
Don't hesitate.
Casey:
Pull over the car.
Casey:
Use a signal.
Casey:
Because you're an adult and courteous, use a signal.
Casey:
If you're walking, do what you need to do to indicate to those behind you.
Casey:
You will be pulling to the side of the sidewalk.
Casey:
If you're biking...
Casey:
God help you.
Casey:
But do what you got to do as a bikist, as Marco would say, to get to the side of the road and not get run over by a car.
Marco:
That makes people so mad, by the way.
Marco:
I know.
Marco:
I know it does.
Marco:
They don't get that it's a joke.
Marco:
It makes them so mad.
Casey:
I know.
Casey:
We know that a bikist is not a thing.
Casey:
Anyways, all kidding aside, go to atp.fm slash store.
Casey:
Check out our merch.
Casey:
Check out our wares.
Casey:
And remember...
Casey:
If you are a member, you get 15% off the ATP store during this time-limited sale and all other time-limited sales.
Casey:
Now is the time.
Casey:
I know we kind of glossed over it last week, but John, would you mind just doing a brief overview of everything?
Casey:
Just one last time to really sell it, just nail it, and send it home.
Casey:
Let's get a few more sales, John.
Casey:
What do we have on offer?
Wow.
John:
We're not going to go through all the merch again.
John:
They know what's there.
John:
People know.
Casey:
All right, I'll go through all the merch.
Casey:
Why got to be such a party pooper?
Casey:
All right, here we go.
Casey:
I'll do a speed run.
Casey:
We've got ATP Pixels.
Casey:
We've got ATP Space Black.
Casey:
We've got the M3 shirt, the M3 Pro shirt, the M3 Max shirt.
Casey:
We've got ATP Six Colors, which is six colors of fabric.
Casey:
Actually, more than six colors of fabric, but the text is always white.
Casey:
Really should have workshopped that name, but here we are.
Casey:
uh atp logo shirt the og we've got the atp hoodie atp polo don't let me down polo ponies polo people don't let me down this is my jam get an atp polo please and thank you we've also got the atp mug in now white and a bluey purpley sort of thing john will correct the actual it's cobalt please casey thank you of course i said that it's definitely cobalt definitely not a bluey purple recycled cobalt don't worry
Casey:
Uh-huh, sure.
Casey:
And the ATP pint glass.
Casey:
So go now, atp.fm slash store.
Casey:
And remember, if you're a member, go to your member page landing strip, whatever you call it, thing, and go get your bespoke discount code for 15% off.
Casey:
atp.fm slash store.
Casey:
This is it, people.
Casey:
The 26th, that's all you got.
John:
The only thing I'll add is remember that a lot of the products come in long sleeve t-shirt versions and sweatshirt versions.
John:
I think this is kind of the first time we've done those in a lot of these products.
John:
So check, even though a lot of times it will show a picture of a t-shirt.
John:
If you watch long enough, the picture will rotate and you'll see some long sleeve varieties.
John:
And it's, you know, the cold winter months in our parts anyway.
John:
So think about those long sleeve options if you're interested.
John:
And this sale does span Black Friday.
John:
We did this intentionally.
John:
So it ends on the 26th.
John:
Black Friday is a couple days before that.
John:
You're going to be thinking about shopping for the holidays.
John:
Probably you'll be bombarded by messages about shopping for the holidays.
John:
Don't forget.
John:
This is your last chance.
John:
I know we do the sale kind of early, but we're trying as best we can to have a chance of getting these things to you in time for the holidays.
John:
So, you know, nerdy gifts for yourself or for others.
John:
Don't forget when Black Friday arrives.
John:
That's another reminder of you to hear Casey's voice in your head is saying, oh, wait, was I supposed to buy some nerd thing?
John:
Yes.
John:
ATP dot FM slash store.
Casey:
Indeed.
Casey:
There's also ute sizes as well, which we added a little bit late.
Casey:
Apologies about that.
Casey:
So you can check out for your kids.
Casey:
You can get them stuff for friends, kids, you know, whatever you want to do.
Casey:
ATP.fm slash store.
Casey:
I'm telling you, this is good stuff.
Casey:
You should definitely check it out.
Casey:
And time's running out, my friends.
Casey:
Time is running out.
Casey:
I feel like that's something else I wanted to add, but I forgot what it was.
Casey:
So that's all right.
Casey:
We'll just make me feel, we'll make me sound smart in post.
Casey:
Right, Marco?
Casey:
Right, right, right.
Casey:
Okay, good.
Marco:
Garbage in, garbage out.
Casey:
Oh, brutal.
Casey:
Brutal.
Casey:
Well played, but very mean.
Casey:
Keyboard wear.
Casey:
Let's talk about... God, you really derailed me with that.
Casey:
Keyboard wear.
Casey:
Let's talk about what happens to keyboards because of your damn dirty fingers.
Casey:
Ryan Holmes writes, Casey, I think what you're seeing as finger grease is likely what keyboard enthusiasts call quote-unquote shine that happens to ABS plastic.
Casey:
It's relatively soft plastic that is susceptible to finger oils and heat.
Casey:
To John's point, some people's finger oils shine them more than others.
Casey:
When you use a magic eraser to remove the shine, you are literally abrading off a layer of plastic.
Casey:
Okay, so, interrupting briefly, I don't debate that that's very possibly true, but I'm not, like, sawing the keyboard with a magic eraser.
Casey:
Like, just a quick glance with it is more than enough, and I only do this a handful of times a year.
Casey:
Well, but, like...
Marco:
that that is true though like that is how the abrasive and a magic eraser works you are taking a layer off it's just a very small layer because it's a very fine abrasive but you're also leaving bits of the magic eraser on your keyboard and going down into the i just i'm still against the magic eraser yeah it's it's not a great option to have to use ideally you don't need it that like because because you're right there is some risk of like you know getting little little dots of it stuck in there and jamming up the keys obviously now that we have a little more robust key
Casey:
boards than we used to that that risk is lower uh but but it is still a risk so yeah you are better off not doing that if you can help it that's fair continuing back to what ryan holmes was writing abs is nice for keycaps because they are easy to double shot meaning the letters can be embedded into the key and won't wear off it's a separate layer of plastic molded inside this also enables any color combo a more durable option is pbt polybutylene tera foot oh my gosh that's all at once
Casey:
which takes much higher temperatures to deform.
Casey:
It does not quote-unquote shine.
Casey:
It often has a deeper sound and can be double-shot, but double-shotting it is not common and can involve a blend with ABS.
Casey:
The more common way to print legends on PBT...
Casey:
is dye sublimation, which involves heat and dye.
Casey:
The dye embeds into the top layer of the key, so the legends will essentially never wear off.
Casey:
There is so much lingo here.
Casey:
My word.
Casey:
I know every, like, interest and profession has their own vocabulary, but my gosh.
Casey:
Anyways, it's difficult to get them as crisp as double shot, however, and they will not let LEDs glow through.
Casey:
PPD is also difficult for light legends on dark caps since the whole cap will need to be dyed except the legend, which is masked out.
Casey:
This amount of dyeing involves longer exposure of the plastic to heat and often deforms larger keys, most notably space bars.
Casey:
Cheaper OEM keyboards use other kinds of surface printing that wears off relatively easily.
Casey:
So that was way more information than certainly I needed or wanted to know about keyboards.
John:
Some background on Apple's keyboards.
John:
So Apple's using ABS, which is the softer plastic.
John:
Apple has used PBT in the past on older laptops before your time.
John:
But back in the day, the power bricks used to come with keyboards that it was kind of hilarious.
John:
They look kind of like, I mean, you've all seen the Apple Extended 2, right?
John:
Even if you've never actually touched one.
John:
What those keycaps look like?
John:
Imagine those, but imagine you took them into a 3D program and you squished them so they're not as high.
John:
right and that's what they made the powerbook keys keycaps look like from the top it looked like oh that's just a regular key but then you look from the side and you'd see they were squished down like they had the same kind of like slanted edges but just much much smaller i'm pretty sure the back in those days the keycaps were pbt of course they weren't backlit so that's the whole big thing i think all of apple's keyboards now on their laptops are backlit you can't backlit like the pbt because it doesn't
John:
light through it that's why they were talking about having abs mixed in because you could let the light go through the abs part like just the letter that's on them or whatever but anyway apple's keycaps on all their modern keyboards all appear to be the uh very much softer abs
Casey:
anonymous writes the backlit keyboard keycaps on apple laptops are injection molded in clear plastic they are painted with white paint then black paint the glyphs are then laser etched using a method that removes the black paint but not the white paint lastly a protective clear coat is applied
Casey:
Andrew writes, your conversation about where on keyboards reminded me of a recent article by a British cycling journalist.
Casey:
Would that be a bikest, Marco?
Casey:
Is that correct?
Casey:
A bikest journalist.
Marco:
A bikalist.
Casey:
Oh, my.
Casey:
We're going to get so much hate mail.
Casey:
On the recent demise of his 10-year-old MacBook, which had survived an impressive amount of abuse, you might like to see how worn his most used keys were.
Casey:
We will put a link in the show notes and maybe Marco will make this chapter art.
Casey:
Maybe, maybe.
Casey:
Let me tell you, actually, you know what?
Casey:
I would not advise you to make this chapter art because this keyboard is freaking horrifying.
Casey:
Not only is this the ridiculous Tetris key return key, which I know the British people love, leaving that monstrosity aside, this is an abomination.
Casey:
Look, this is disgusting, this thing.
Casey:
I can't handle it.
Casey:
I need to scroll down.
John:
This is one of many photos that we got.
John:
This is the worst one, obviously.
John:
But this is one of many, many, many photos that we got, most of which were not 10 years old, to be clear, most of which were much younger laptops.
John:
And to visualize what this picture looks like, picture an Apple keyboard with all of its black keycaps and then have little circles worn through them at various places.
John:
Seeing the mechanism beneath because as the earlier feedback item said Apple's keys are made from clear plastic and then they have black paint and white paint over them and so this this imagine the keys with that with the black and white paint rubbed off and
John:
just showing you the clear plastic.
John:
It's like these little window panes of various sizes.
John:
And you can see where they wore through the black first and then the white second, right?
John:
And you see kind of like a terrace pattern where it's like black, then there's a white ring, and then there's a clear spot.
John:
And you can tell where this person hit the space bar, like in one very specific spot all the time.
John:
Yeah, tons of people had this.
John:
And, you know, again, the leading theory is...
John:
ph levels either more acidic or more basic uh you know sweat from their fingers or whatever essentially wearing through the paint because these are painted uh clear plastic and if they were pbt where it was the the the color of the plastic instead of being painted this wouldn't happen but then you couldn't have the backlights unless you double shot them but then there's the you know they're saying like to put the die on the entire key or something a long key like the space bar might warp so i would say like as we mentioned in the last show um
John:
keyboard durability on laptops uh aesthetic uh durability not like you know making the keys continue to work still seems like an area where apple could use some improvement there i think what they have now is a reasonable ish compromise because they're lightweight they they're fairly sturdy they hold up pretty well most of the time but if you're one of those people who wears away keycaps it's a bummer for you because you're going to have a worse experience with this keyboard and
John:
It'll be nice if the keys still work, and I guess as long as you remember what those letters were before you wore them off with your fingers, you're fine.
John:
So functionality is job one, but aesthetics should be somewhere in the top five, and the current keyboards are failing that for some subset of Apple's customers.
Casey:
This next section is John tries to make Casey feel bad about liking 14 inch laptops.
Casey:
So I'm just going to.
Marco:
It's not just John.
Marco:
I mean, if only one of us would have said last episode, you know, with this amount of power draw, so much increase from the M1 generation, maybe the 14 inch might be a bad idea if you get the max chip, because it seems like that's a lot of heat for the 14 inch to quietly and gracefully get rid of compared to the 16 inch.
John:
So, John, what happened?
John:
I mean, to be clear, this was true of the M2 and M1 generations as well, which is why Marco gave that advice.
John:
But anyway, we'll put a link in the show notes to a YouTube video where they tested the M3 Max 14 inch versus 16 inch.
John:
And I think I said in the last show that like, well, you know, maybe the cooling should be roughly equivalent.
John:
And the thing that is stuck on top of the M3 Max, like the little whatever heat spreader thing that's pulling the heat off.
John:
Yeah, that part's the same.
John:
Every other part of it is not the same, though.
John:
First of all, the heat pipe itself is leading away from the SoC.
John:
That heat pipe is wider on the 16 inch.
John:
So right away, there's a difference in the cooling system.
John:
But the big difference is the fans in the 16 inch are so much bigger than
John:
fans in the 14 inch like they're really using that extra space to good effect uh it is a dramatic difference in the size of the fans and obviously the amount of air they can move then the fans are blowing that air over uh you know these little fin things that are attached to the heat pipe and there's more of them uh anyway in the testing uh two big things came out first that 14 inch is going to run the fans at much higher rpm so they did like
John:
Cinebench benchmark and the 14-inch was running the fans at 7,200 RPM and the 16-inch at 3,800.
John:
So it was like double the fan speed.
John:
And obviously, if the fans are running faster, they're going to be noisier.
John:
And since the 14-inch fans are smaller, they're going to be a little bit more annoying too because they're higher pitched.
John:
That's not great.
John:
They stabilize fan speed.
John:
I love when they fancy up their benchmarks here.
John:
So the average speed in the Citebench 2024 GPU test was 7,200 RPMs for the 14-inch, 1,700 RPM for the 16-inch.
John:
That's...
John:
big difference that's that's as in like you can hear the fan screaming the entire time um in the 14 inch and on 16 you probably wouldn't hear anything and now casey may be saying i don't care about that i don't run 3d benchmarks i don't i don't care about uh the the average fan speed over a 10 minute run of some type of thing but then they did an xcode test and that's really really hitting that casey where it hurts
John:
Because that's the thing that he does do.
John:
And this is not about the fan speed or about temperatures or anything like that.
John:
This is the fact that all those things with the fan being higher on the 14 inch translate to the 14 inch being in more thermal distress.
John:
And thermal distress means thermal throttling.
Marco:
and so this is where the rubber meets the road the 14 inch is a little bit slower than the 16 inch because it thermal throttles when doing xcode compiles now this is probably a big xcode compile maybe it's bigger than any of casey's apps maybe it doesn't affect him or whatever and the difference isn't huge yeah because this is it took like almost a minute and a half to do this compile so like if you're if you're doing a very long sustained cpu or gpu uh drain you're gonna you're gonna hit throttling on the 14 inch no question like that's what this is clearly showing like
Marco:
If you are pegging the CPU or GPU for a minute or more, you are almost certainly going to see throttling on the 14-inch that is not present on the 16-inch.
Marco:
yeah and the xcode benchmark was 82 seconds on the 14 inch and 72 on the 16 inch so not a huge difference not like the fan speed difference i feel like the big quality of life difference is the fan speeds like i mean that's like an m2 to m3 difference in performance though like but again like this is a this is a long sustained one like you know if you know my build takes 11 seconds like it's probably in in that span i probably wouldn't see the difference between a 14 and 16 but if you're doing builds to take a minute you probably will
Marco:
Yeah.
John:
So, I mean, as the case with all these laptops, right, it's just a question of how much the cooling system can fight off the inevitable heat saturation due to long-range jobs.
John:
It's rare that a laptop, a very high-powered, top-of-the-line, stuffed-to-the-gills laptop, can sustain its full power for a very long period of time because the cooling systems usually aren't adequate to get all that heat out of there to keep up with it.
John:
Especially if you have some contrived, like...
John:
CPU maxed out at the same time as GPU maxed out at the same time as neural engine max out at the same time as a video encoder matched out.
John:
Like they're not going to be able to deal with that.
John:
But almost no real job is like that.
John:
That's why the GPU benchmark is so brutal here because this was just a benchmark of the GPU and the fan speeds were 7000 RPM versus 1700.
John:
That's not even close.
John:
That's just they don't even just stressing the GPU is
John:
is enough to really really see a difference and that's relevant people playing games if you're going to play games on your laptop you're like i want to you know one of these m series laptops i don't need to hear the fans screaming when i play games get a 16 inch uh you'll you have a fighting chance of making that real because if you you know if you want to find something that's going to really stress the gpu yeah playing a game for a long period of time will do it um
John:
I don't think this means that a KC should get a 16-inch.
John:
I think the 16-inch is too big as well.
John:
I think the whole point of being portable is to have something small.
John:
But if you are concerned about sustained performance, A, don't get a laptop, but B... Hey, the 16-inch sustains it just fine.
John:
Well, no, but the 16-inch eventually heat soaks as well, especially if you do a CPU plus GPU combo.
John:
These benchmarks are a little bit shorter.
John:
yeah and also the other thing they were measuring which isn't any of these graphs in the show notes here is clock speed like they're just seeing like what is what are they hitting their rated top clock speed and both of these laptops occasionally depending on the benchmark would never actually achieve their their rated top benchmark top clock speed or would do it very briefly in the beginning of the test and then never see it again right so there's always room for improvement right 16 inches versus 14 doesn't mean the 16 inch never thermal thralls it just means it does it less than the 14 which is why it's coming out ahead in these tests
Marco:
I would say these tests prove my hunch last week that if you're going for the Max chip, maybe don't get the 14-inch.
Marco:
If you're getting the 14-inch for an all-around balance of portability and size, and maybe you're going for the M3 Pro chip, that's a different story.
Marco:
It can probably handle that a lot better.
Marco:
But if you're getting the M3 Max...
Marco:
and you intend to push it at all, which you probably are if you're getting the Max, you should be aware of this.
Marco:
This might not change your mind, but this is something that you should really be aware of, that the M3 Max seems to be substantially crushing the 14-inch thermal system more than the M1 generation did.
Marco:
The M2 was probably somewhere in the middle, because its power was between the two.
Marco:
But they've ramped up the peak power draw of these chips compared to previous generations.
Marco:
The M1 draws less power than the M2.
Marco:
The M2 draws less power than the M3 Max in all those cases.
Marco:
And so be aware of that, that they have raised the power envelope without increasing the thermal dissipation capacity significantly.
Marco:
So if you're coming from a 14-inch M1 Max and you're thinking, well, it handled my M1 Max just fine in the 14...
Marco:
The same can't necessarily apply to the M3 Max.
Marco:
You have to reevaluate because they're not drop-in replacements in terms of power and thermals.
Marco:
The M3 Max uses substantially more power and makes substantially more heat under load than the M1 Max did.
Marco:
So that might change your calculus of whether you want a 14 or a 16.
John:
i think the cooling the cooling system is upgraded a tiny bit in the m3 generation versus the m2 like i think the heat spreader thingy on the soc is a little bit larger in this generation and i'm not sure about the fans if they are exactly the same um but i i don't i don't think it's entirely true they just use exactly the same cooling system otherwise it would never be able to do it because the the power difference and the 16 inch comparing the m2 16 inch versus m3 max 16 inch
Marco:
it's pretty big difference the m3 max is pushing like 50 watts and the m2 max was only ever getting like 36 or so so there's a lot of extra heat yeah and the m1 generation i think was in the 20s like it's it's substantially more power and heat here and also that that also results in a larger difference between the two models in battery life than there used to be there was always a decent difference because the 16 inch battery is just much bigger um and yeah the screen uses more power than the 14 it's bigger but not that much more power and
Marco:
So the 16 is still the battery monster.
Marco:
But in this case, keep in mind, again, the CPU is drawing more power than the previous generations at the same marketing name level.
Marco:
And also, I don't know how much of a factor this is, but if those fans are spinning faster...
Marco:
Fans also use power.
Marco:
So the result after these tests, the result in this video, the 14-inch battery was substantially more drained than the 16-inch.
Marco:
It was not a small difference.
Marco:
So again, just don't make assumptions based on whatever worked for you in the M1 generation.
Marco:
Don't assume that the same thing will work for you in the M3 generation because the power levels are very different.
Casey:
I understand everything that you guys are saying, but I...
Casey:
I don't feel like I'm bumping into thermal throttling issues any time other than when I'm doing transcosals with FMPEG, at which case, okay, yeah, I'm reaping what I sowed or whatever the turn of phrase is, you know what I'm trying to say.
Casey:
Okay, that's fine.
Casey:
The other thing is I didn't want to go down from 64 gigs of memory.
Casey:
I probably could have and been fine, but I didn't want to.
Casey:
And in order to do that, in this generation, you have to have an M3 Max.
Casey:
Like, that's the end of the meeting.
Casey:
So in fact, you not only need an M3 Max, you need the baller M3 Max.
Casey:
That's the only, I'm looking at the configurator right now, it's the only option in the 14.
Casey:
So if I want 64 gigs, well, guess what?
Casey:
I'm getting an M3 Max.
Casey:
In fact, the Pro appears to top out at 36 gigs.
Casey:
which even for me is probably fine.
Casey:
But I didn't want that.
Casey:
I wanted 64.
Casey:
And honestly, I have zero regrets about this machine.
Casey:
I've been using it at home.
Casey:
I've been using it away from home.
Casey:
I freaking love this computer.
Casey:
And I mean, I freaking love my M1 Max as well.
Casey:
Don't get me wrong.
Casey:
But this computer is great.
Casey:
I haven't heard the fans except one time when I was doing a transcode.
Casey:
I have no reason to believe battery life is suffering compared to the last one.
Casey:
It's been flawless so far.
Casey:
I don't argue any of what you guys are saying.
Casey:
I'm not trying to say that you're wrong or lying or anything like that.
Casey:
It's just in real world use, I am not personally noticing, not to say it's not affecting me, but I'm not noticing the results of any of these problems in the 14.
Casey:
And for me, and I totally understand, Marco, why you reached a different conclusion, but for me, I don't want a frigging aircraft carrier in my backpack.
Casey:
And I travel with this machine often enough to
Casey:
that that matters to me.
Casey:
And so I made the choice I made, and I think it's the right one for me.
Casey:
You made the choice you made.
Casey:
I think it's the right one for you.
Casey:
And everyone's happy.
Casey:
Why won't you just let me be happy?
Marco:
No, look, and I think it's important to point out, like, if the 14-inch still worked for you, that's great.
Marco:
But...
Marco:
People who are buying it should know this ahead of time, that it's going to have these trade-offs, that the M1 generation of it had fewer trade-offs compared to the M3 generation.
Marco:
And if what you are seeking is the highest-performing laptop...
Marco:
In the M1 generation, you could say, well, these are basically the same.
Marco:
Now, you can't really say that.
Marco:
Now, the 16-inch is noticeably higher performing with a lot of workloads.
Marco:
So, it is worth knowing.
Marco:
Also, again, if stuff like fan noise is really critical to you, then, again, that's something that's worth knowing ahead of time.
Marco:
So, that's why.
Marco:
And for many people, they're going to still get the 14-inch like you and be totally happy with, and that's great.
Marco:
But, again, you've got to know going into it...
Marco:
The assumptions that you might have made from previous generations of the M-series laptops are a little bit different now.
John:
I do wonder if there's not a low-tech solution to this.
John:
Remember we were talking about a little RGB fan cooler thing on the back of your phone to get sustained frame rates?
John:
Just get some giant plate.
John:
Yeah, like I know they make laptop cooler type things, but I do wonder... They're really quiet and elegant.
John:
Yeah, right.
John:
But I do wonder if you say you get a 14-inch and you want to use it in desktop mode and you don't want it to thermal throttle, just how much extra cooling would it take?
John:
probably not that much like and probably could be applied externally i don't you know again caveats about condensation and don't destroy your computer and blah blah blah whatever but like based on how well it works on the little phone thing to sort of sustain frame rates when the phone could not do it without the cooler i do wonder if just a
John:
little bit of extra help on a 14 inch would help it uh not thermal throttle at all because it's you know it's the fans are almost doing the job because yeah the fan speed difference is huge those fans are working overtime right but the benchmark difference is not that big so it's just it's like the the 14 inches uh cooling system is overmatched but just by a little bit so i wonder if you just gave it a little bit of help um which is unlike lots of other computer things like oh i didn't get enough cores well there's not really anything you can do to add more cores to your computer
John:
But if thermal throttling is your problem, pointing a fan at the back of your laptop when it's like sitting upright, you know, like in a bookshelf type thing might be enough.
Marco:
One potential limitation there is, remember we learned this from people who wrote in like a couple of years ago, this came up where we were talking about like the MacBook Air thermal design without any fan, that like there are certain standards that different, you know, governments and things set about how hot the outside of a laptop can get for safety reasons.
Marco:
And so the way these are designed, usually the the processor is not like heat pipe directly to the exterior case.
Marco:
Like there's usually some kind of air gap there so that the processor is intentionally not transferring its heat much to the metal case.
Marco:
And because of that.
Marco:
It actually makes it difficult to make external coolers that are very effective because you're not getting all that heat away from the chip through the case very much.
Marco:
It's mostly dumping it into the air.
Marco:
So it's a little bit harder.
Marco:
So you couldn't, for instance, just stick like a Peltier element cooler on the back of a laptop and have it be...
John:
Very good.
John:
You just need a temperature differential.
John:
And one exists.
John:
I mean, feel the bottom of a laptop when it's hardworking.
John:
I know what you're saying.
John:
Not all the heat is being dumped out there, but there's still a temperature differential.
John:
That's a test that someone should try.
John:
Just consider that.
John:
I guess the reverse of it is like...
John:
uh if you're doing some sustained workload with your laptop don't throw it on your bed on top of your comforter like don't don't you know what i mean like yeah you see how kids treat laptops they have no idea that these are like living breathing things right it's a little empathy for the machine here don't take your laptop and just put it on a big soft pillow surrounded by blankets like it needs you need to get it with the lid closed of course right naturally give it give it a fighting chance
Casey:
Yeah, and just as a final note, because I'd like to move on, I feel like there's a lot of FUD going around right now.
Casey:
I don't have any anecdata that indicates that this problem is any worse with this generation than it was with my two-year-old laptop.
Casey:
Like, yes, again, I'm not trying to argue that the facts that you're presenting to me are wrong.
Casey:
All I'm trying to say is that in day-to-day use, I straight up do not hear the fans.
Casey:
So, yes...
Casey:
In the occasion that I can hear the fans, yes, they probably are faster, they probably are more annoying, they're probably louder.
Casey:
I don't debate that, but I feel like we're making a problem out of nothing here.
Casey:
This machine is phenomenal.
Casey:
If you want a desktop replacement that you occasionally move, 100%, get the 16, no argument.
Casey:
But if you want a computer that you move occasionally, and you favor the ease of moving it over anything else, which is where I am,
Casey:
Get the 14.
Casey:
Don't let these two knuckleheads try to convince you otherwise.
Casey:
It's a perfectly good computer.
Casey:
There's no need to... I'll try to convince.
John:
I also agree.
John:
I think the 16 is just too big for a laptop.
John:
And really, you should just get a desktop anyway.
Marco:
Oh, my God.
Marco:
It's so good.
Marco:
It's so good.
Marco:
The 16 is so good.
Marco:
There's basically no trade-off except...
Marco:
It is large.
Marco:
But again, when you look at... We are so spoiled now.
Marco:
The 16-inch MacBook Pro is lighter than most 15 inches of most of our computing-using careers.
Marco:
And it is 16 only because they made the bezel smaller.
Marco:
So the footprint is roughly 15-inch glass that we've had forever.
Marco:
The weight is similar or lighter than most 15-inch glass laptops we've ever used.
Marco:
It is glorious.
Marco:
I'm not going to argue that the 14-inch isn't noticeably smaller and lighter.
Marco:
It is noticeably smaller and lighter.
Marco:
I would maybe argue that...
Marco:
If that's really your top priority, maybe consider the Air.
Marco:
But the 16-inch, again, it's all relative.
Marco:
Yes, it is larger and heavier than yours, but compared to anything ever in the history of computing, these are all very thin and light, and it's fine.
Marco:
And even, again, when you're looking at... Suppose you're putting it in a backpack and bringing it to wherever.
Marco:
The weight of your backpack, empty...
Marco:
is probably a few pounds like it depends on what kind of bag you have the weight of like your total carry of what what most people bring around in their backpacks every day maybe 10 or 15 pounds and so the difference in laptop weight of like a one pound difference between two things that's fair is not really that significant it feels significant when it's in your hand
Marco:
But however you are carrying it, generally speaking, for most people, the difference between a 16-inch MacBook Pro and a 14-inch MacBook Pro being like a pound or whatever is not a massive difference in whatever they're carrying it in total.
Marco:
So again, if it matters to you and if you really still want to minimize it, great.
Marco:
I get that.
Marco:
Please do.
Marco:
But...
Marco:
These are trade-offs that you should be aware of.
Marco:
Despite what you said, Casey, they weren't as severe of trade-offs in the M1 generation because the total power usage of the M1 chips was way lower.
Marco:
Rather, the peak power usage, like under load, the peak power usage was way lower in the M1 generation than it is in the M3 generation.
Marco:
So that is a difference that is worth noting.
Marco:
And before, again, with the M1 generation...
Marco:
You could say these computers are basically the exact same computer, just bigger and smaller.
Marco:
And now that's a little harder to say.
Marco:
There's a much bigger asterisk on that now.
Casey:
So let's talk about chip packaging.
Casey:
Johnny Sruji had an interview in July with his alma mater, who is a college.
Casey:
The name is escaping me.
Casey:
It was somewhere in Israel.
Casey:
And he went back there and did an interview with somebody there and talked a bit about chip packaging.
Casey:
John, do you want to either intro this or read it to me or tell me how you want to handle this?
John:
Yeah, so this is the interview that I referenced on a past show where I forget what we're talking about.
John:
We're talking about three nanometer processes or stuff like that.
John:
But anyway, I've said that Apple executives won't reveal anything about future products, obviously, but anytime one of them talks, they'll usually say something that lets you know some vague future direction.
John:
And in the interview, Johnny Surgey said that he thought that packaging was a potential area of innovation.
John:
And I wanted to follow up on that today to first provide the excerpt from the interview because the whole blah, blah, blah is interesting is I think that was like Tim Cook saying the wrist is interesting or some BS like that.
John:
And whatever the quote was from Jobs when he was talking about cell phones was similar.
John:
That's how we all know Apple is making a phone.
John:
So I just wanted to get the actual words that he said about packaging because there was also recently an interview with
John:
uh of uh intel ceo pat gelsinger with uh ben thompson of uh stratechery or stratechery however you want to pronounce it um where uh yeah where the intel ceo talked about you guessed it packaging and how important it is to intel and for the future and so i feel like this is a a confluence of events that's worth digging into deeper even if this as we said in past episode of this generation it just looks like the m3 is going to be
John:
M3 Ultra is just going to be two M3 Maxes stuck together just like it was before, which is fine and great and will be awesome.
John:
But keep your eye on the future for packaging stuff.
John:
So, yeah, Casey, you just want to go through the interview segment.
Casey:
Sure.
Casey:
So how about I'll play the role of the interviewer and you can play the role of Johnny Sergi.
John:
Sure, give me the longer one.
Casey:
All right, fine.
Casey:
You know what?
Casey:
I'll take one.
John:
I don't know if we can play acted.
John:
I'm going to get Marco to be mixed in.
John:
He can be the announcer.
John:
Oh, gosh.
Casey:
All right, so you be the interviewer.
Casey:
I will be Johnny.
Casey:
Let's do it that way.
John:
You don't need two people to read them.
John:
You can just be one person.
Casey:
Play with me in the space, John.
Casey:
Come on.
Casey:
Are you really just going to leave me hanging like this?
Casey:
Jesus Christ.
Casey:
Is it my line?
Casey:
Yes, it's you.
Casey:
We're actually doing this?
Casey:
Yes.
Casey:
Yes.
Casey:
You are the interviewer, John.
John:
All right.
John:
This is the interviewer.
John:
What are the next challenges in the design and architecture of processors that Apple should tackle to get to the real next generation of processing?
John:
This is where the interviewer starts asking questions like they shouldn't ask.
John:
What's the next challenge?
John:
You know, what should they tackle to get to the real next generation?
John:
I'm not asking a question about the future, but I totally am.
John:
Anyway, here's what Johnny Surgey said.
Casey:
It is getting more and more challenging.
Casey:
Those of you who follow CMOS technology, whether it's 5 nanometer or beyond, that's getting harder and harder, which I think is great, by the way, because if it gets harder, that means Apple will try better and better.
Casey:
So that's good.
Casey:
I'm not going to describe our future roadmap, but there are many challenges.
Casey:
For example, when you take CMOS technology, I think one of the things that is going to be important is packaging without getting into details.
Casey:
So the way you package the chip is going to be important, or maybe you architect the chip in a different way.
John:
and i know you think like that's saying nothing what are you talking about this was the quote yeah they don't say a lot but the fact that they say anything about anything oh packaging just just that word packaging what does that mean i mean again apple already does interesting things with packaging the silicon interposer is an interesting thing done with packaging and we've talked a lot about packaging technology in the past but this is the tiny tidbit uh they left you know maybe you architect your chip a different way packaging is interesting uh i do like these typical apple uh
John:
bravado uh where they said oh you know the the silicon the world of silicon is it's getting more and more difficult and that's great for apple because when things get hard like we will excel like that'll make us try harder because we're the best and basically bring it on because the harder it gets the more you'll see that apple is the best which
John:
which is great for him to say.
John:
We'll see if it actually turns out to be true, how those cell modems going, Johnny.
John:
I thought that was some top tier Apple executive bragging in this interview.
John:
But anyway, that's what he said about packaging.
John:
He just dropped the word packaging.
John:
So now here is the Intel CEO, Pat Gelsinger, also talking about packaging.
Casey:
So I will be playing the role of Pat Gelsinger in this one since that's the longer one.
Casey:
Okay.
Casey:
Gelsinger says, this idea of chiplets, I think, is the new way that all chips get designed.
Casey:
The idea of advanced packaging multiple chips into the advanced package and whether that's an MCP or multi-chip packaging, whether that's a two and a half or 3D construction, I do think that becomes the standard.
John:
And this is Ben Thompson.
John:
The interviewer saying, why is advanced packaging the future?
John:
You can see you can ask Intel questions about the future.
John:
I know this has been a big focus for Intel.
John:
It's something that you want to talk about.
John:
And from everything I know, your technology is leading the way.
John:
Why is that so important in addition to the traditional Moore's law?
John:
Why do we need to start stacking these chiplets?
John:
Give me the top reasons.
John:
So this, you know, Ben is asking a good question here, which is like.
John:
Of all the things, the problems Intel has, and you can read the full interview, you know, he's interviewed the new Intel CEO a few times.
John:
Intel's got some problems.
John:
They fell behind.
John:
They're, you know, getting beat in the fab business by TSMC.
John:
Why in the world is Intel caring about packaging?
John:
And more importantly, why is Intel putting tons of money into innovations in packaging?
John:
And so he's asking, like, you know, what's the deal?
John:
Give me the top reasons for the whole focus on packaging.
Yeah.
Casey:
One is, now you're able to take the performance-sensitive transistors and move them to the leading-edge node, but leverage some other technologies for other things.
Casey:
So you get to mix and match technologies more efficiently, effectively this way.
Casey:
John, would you like to jump in and explain what I, Pat Gelsinger, just said?
John:
Yeah, so this is part of Intel's problem, is they, you know...
John:
their fabs fell behind.
John:
They couldn't get off of whatever... Or were they stuck on 14 nanometer for ages, right?
John:
And TSMC went ahead of them and has been ahead since then.
John:
And so they're like, look, we know at Intel that we have some fab problems.
John:
In fact, Intel is paying TSMC to fab a bunch of its stuff.
John:
And they're saying, one of the things we can do by breaking the chips up into smaller individual pieces...
John:
is get the most important parts to be fabbed by TSMC, not by us, on the good process node, and then use other lesser fabs, like Intel's fabs, to fab the parts of the chip where it doesn't matter as much.
John:
So if you're Intel, you don't want to pay TSMC, and Intel, you can read all about it in Ben's coverage, but Intel is trying to split where they're like, we're going to have a fabbing part of the company, and then we're going to have the chip design part of the company, and we're going to pretend like they're separate.
John:
So the fabbing part is way behind and needs to catch up, but the part that owns like x86 and makes the chips, we think that's good.
John:
So, you know, they're using TSMC, but they don't want to have US TSMC for everything.
John:
That would get very expensive.
John:
Intel's big advantage is they...
John:
Fab their own chips and they design their own chips, their, you know, vertical integration and everything.
John:
So like this is why one reason why packaging is important.
John:
Intel can pay the minimum amount required to the company with the good fabs to make the part of the chip where it's most important for it to be on the top process level.
John:
And then they can use cheaper stuff for the other parts.
Yeah.
Casey:
Second, we can actually compose the chiplets to more appropriate die sizes to maximize defect density as well.
Casey:
If you have a monster server die, you're going to be dictated to be n-2, n-3, just because of the monster die size.
Casey:
I get to carve up that server chip.
Casey:
I get to move the advanced nodes for computing more rapidly and not be subject to some of the issues.
Casey:
Defect density early in the life of a new process technology.
John:
So this is about Intel setting aside SOCs, where it's a whole bunch of stuff on a big giant.
John:
It's a system on a chip.
John:
They're giant server chips, like the huge 56-core Xeon things, which are just a single gigantic chip.
John:
And you have to fab that.
John:
It's fabbed in a single die.
John:
and if some part of that screws up like the more area you start covering on the silicon wafer the more stuff you put there the more chances there's going to be some error and you might have to throw out the whole chip and so this is another advantage to packaging is like if you just if you just make it out of smaller parts a if it's bad you're only throwing away like a smaller area of your wafer you don't have to throw away this huge square you throw out the little square and b uh your defect density can be lower he says maximizing defect density but whatever he meant minimizing
John:
That you'll get more of those littler things that you fab will come offline and be completely error free.
John:
Whereas trying to just get a single error free humongous Xeon is difficult.
Casey:
SRAMs in particular, SRAM scaling will become a bigger and bigger issue going forward.
Casey:
So I actually don't get benefit by moving a lot of my cache to the next generation node like I do for logic power and performance.
Casey:
I actually want to have a 3D construct where I have lots of cache and a base die and put the advanced computing on top of it into a 3D sandwich.
Casey:
And now you get the best of a cache architecture and the best of the next generation of Moore's Law.
Casey:
So it actually creates a much more effective architectural model in the future.
Casey:
Additionally, generally, you're struggling with the power performance and speed of light between chips.
John:
yes this whole uh technique this is a particular technology that intel has advanced with which lets them take power and feed it through the layers of the sandwich in a more efficient way to get to precisely where they want it i think that's what they're referring to here the sandwiching isn't just because of that but the intel has it has an innovative way to send power up to the bottom of the chip instead of sending through the top because when you send it through the top you got to weave it through around a bunch of obstacles or you send it through the bottom there's nothing in the way because you make these layered sandwiches or whatever that's what that's talking about and and you know sram is like look
John:
The SRAM is what we were talking about before with the GPU RAM that was being shared better with dynamic caching.
John:
It's faster than the RAM that's out under RAM chips, like your main memory, but it's not the same as registers.
John:
It's kind of in between there.
John:
It's like very expensive RAM, or each little bit of RAM takes a large number of transistors, but it's much faster than regular RAM, and it's in the CPU.
John:
And I say, I'm like, you don't get any benefit of doing that on three nanometers.
John:
So not only do you not want to waste money doing it because it would be more expensive, you don't get benefit from doing it anyway.
John:
So keep that on a different lower process and then sandwich it all together.
Casey:
Any other questions for me, Ben?
John:
Yeah, I'm Ben again.
John:
So how do you solve that with chiplets when they're no longer on the same die?
Casey:
In the chiplet construct, we're going to be able to put tens of thousands of bond connections between different chiplets inside of an advanced package.
Casey:
So you're going to be able to have very high bandwidth, low latency, low power consumption interfaces between chiplets.
Casey:
It also becomes very economical for design cycles.
Casey:
Hey, I can design a chiplet with this IO and use it for multiple generations.
John:
And you can see how all the stuff that I described would be appealing to Apple.
John:
Apple's already making SoCs, which are essentially single dies, system-on-the-chips that have a bunch of stuff in them.
John:
And they're getting pretty big.
John:
That's part of the reason the Ultras have been two maxes weaved together with the silicon interposer rather than just being a single large chip the size of an Ultra.
John:
Because that would be even more expensive to manufacture.
John:
It's also part of the reason we didn't have the quad, right?
John:
So Apple would love the idea that you could take subunits of this, like the thing that does IO or whatever, and reuse that between generations.
John:
Because between generations, like, oh, it's right overrun now, Thunderbolt 3, Thunderbolt 4, that lasts a couple generations.
John:
I think both the M2 and the M3 had the same speed Thunderbolt interface.
John:
So if that was in a separate chiplet, you could just reuse that one from before on the future chip.
John:
That could still be 7 nanometer, for example.
John:
I don't know if this is the right part of the chip to be reused.
John:
You can design it once and reuse that little sub-chiplet in future generations as long as you can do an interconnect between all of these that has all the attributes that they subscribe, low latency, blah, blah, blah.
John:
So far, Apple's packaging innovation has been the Interposer, which is technically impressive and it makes the Ultra possible.
John:
But this is more like a more general purpose thing of like...
John:
lots of different little you know a city of chips lots of different little islands on your chip to even to just make what apple does currently in a single soc and that can make things less expensive and improve your yields as well
Casey:
And then you want to tell me about the Universal Chiplet Interconnect Express.
Casey:
Is that a train, John?
John:
UCI and then lowercase e. So this is one of these multi-company consortiums to do exactly what we were just describing.
John:
So this is a press release from March 2nd, 2022.
John:
It says, Intel, along with Advanced Semiconductor Engineering, AMD, ARM, Google Cloud, Meta, Microsoft, Qualcomm, Samsung, and TSMC,
John:
have announced the establishment of an industry consortium to promote an open die-to-die interconnect standard called Universal Chiplet Interconnect Express, or UCIE.
John:
The chiplet ecosystem created by UCIE is a critical step in the creation of a unified standard for interoperable chiplets, which will ultimately allow for next generation of technological innovations.
John:
Apple's not on that list, but TSMC is, and Apple has a good relationship with them.
John:
So if we're looking towards a future where it becomes untenable to continue making larger and larger dyes, you know, with everything on them and these SOCs, a good industry standard system for making chiplets at whatever the most economical and useful size is and weaving them together to essentially form an SOC out of it.
John:
Seems like a good idea.
John:
Doing it all in one die probably still has advantages, but price may not be one of those advantages.
John:
And again, Intel is super interested because they essentially can't do that without having TSMC do all of their fabbing for them if they want the top tier technology.
John:
Apple doesn't currently have that problem, but the fact that Johnny Surugi mentioned stuff about packaging and the fact that this whole...
John:
chiplets thing like other companies are already doing this amd has been big into the chiplets thing for a while intel is putting a lot of money and research into this and their their chips they're going to be coming out with use some of this technology i think it's unavoidable that apple will start using it too why am i interested in all this you know why it's the keep the dream alive of a mac pro chip that is that is not the same as one you can get in a mac studio and what's going to make that possible apparently with the approach they use for the past two generations not economically feasible
John:
If it's ever going to be economically feasible, we need a change.
John:
This change in packaging could bring that about.
John:
Stay tuned.
John:
I don't know.
John:
2027.
John:
Who knows?
Casey:
Oh, my.
Casey:
All right, John.
Casey:
In keeping with the idea of keeping the dream alive, tell me about Air Network Change.
John:
This is a nightmare.
John:
It's not a dream.
John:
Touche.
John:
Error network change continues.
John:
Some more research from the internet about this.
John:
So here is a Stack Overflow question.
John:
This person asks, whenever my iPhone and Mac OS are on the same Wi-Fi, Chrome and Mac OS often reports the error network changed.
John:
I found that whenever my iPhone and Mac OS are on the same Wi-Fi, a record will often appear in the routing table in Mac OS and disappear after a few seconds.
John:
When this route record appears, my Chrome will most likely have an error network changed error.
John:
I turned off the iPhone's Wi-Fi, the routing and the record in macOS disappeared, and Corona no longer had the error.
John:
So everyone is always looking for, like, what causes this?
John:
And so they start looking at anything, like, what changes my network?
John:
What is it about my network that has changed?
John:
And they'll spot something, and they'll find a thing to attribute it to, and they'll stop that thing by saying, it looks like it's happening because my phone is here, and if I turn off my Wi-Fi, I fix the problem.
John:
Every single one of these that has been reported I have experimented with,
John:
And it doesn't actually eliminate the problem.
John:
It just moves it around.
John:
Because the problem is the network changed and Chrome flips out about it.
John:
Is the problem that the network is changing?
John:
Is the problem that Chrome's flipping out about it?
John:
Is it both?
John:
Either way, if you stop the thing that changes in your network, it doesn't mean that something else won't also change your network and cause the error.
John:
But I'm glad that somebody found something that helps them.
John:
This next one is about Tailscale.
John:
Gil Penderson says the Tailscale VPN used to trigger this, but they fixed it with, and there's a link to a patch.
John:
So this is Tailscale patching their own code to avoid this error because they were seeing this error happening and they were saying, hey, this thing that we're doing, macOS is flipping out about it, so...
John:
you know basically it says ios slash mac os so it's apparently not just a mac thing will reconfigure their routing anytime anything minor changes in the net map and so they're changing their code to not do that because they don't want to anger ios slash mac os this is a change that committed in 2020 so is our ios and mac os you know changing their routing more is there something that was part of the operating system that's causing it
John:
lots of theories about this one that's not even in the notes here is about um if you have a dev device like if you have an iphone that is configured as a dev device there's apparently a new way that max communicate to the phones for purposes of development that also uses a you know network connection that's brought up and torn down and apparently causes network to be changed as far as chrome concerned and it flips out so anyway here's an anonymous bit that's getting us closer to the root of the problem
John:
This is an explanation of it from the perspective of Chrome, Chromium, whatever.
John:
Because again, this is not just a Mac or iOS problem.
John:
This happens to be called Linux.
John:
This is a...
John:
Part of the problem is clearly with the Chrome slash Chromium browser engine, but the operating systems participate in it as well because they're the ones that control the network that is changing.
John:
Anyway, Anonymous writes, the network layer from Chromium is available as a standalone cross-platform library called Chronet, which is open source and used in other non-Google applications.
John:
I work at Google on a major Android app that uses Chronet extensively, so I've got experience with their network changed.
John:
This issue appears to be with QUIC, QUIC, which is the UDP-based web protocol thing.
John:
We'll put a link to the Wikipedia page.
John:
You can read about it.
John:
And or HTTP3, which is a new version of HTTP, which is sensitive to changing networks.
John:
Whenever the networks on the device change, the connection needs to be reset.
John:
This isn't exactly a bug.
John:
It's a necessary step.
John:
Not doing this would mean connections wouldn't work properly.
John:
Unfortunately, this causes error network change when the device's network changed.
John:
As previously mentioned, this can be for non-obvious reasons such as Docker Desktop customizing the networks repeatedly, VPN apps misbehaving, cell signal dropping in and out, etc.
John:
There are ways in which this can be fixed.
John:
First, not using HTTP slash quick is the easiest.
John:
Safari and other Apple stuff doesn't use this yet, so isn't currently experiencing these issues.
John:
However, this will likely change as Apple rolls out HTTP 3 support.
John:
Apart from this issue, it's a much better technology, particularly for mobile devices.
John:
So this is the first offered explanation of, hey, why doesn't Safari have this problem?
John:
The theory is, oh, Safari doesn't use quick or HTTP 3, and so it's not going to see any of these errors.
John:
difficult to test because as far as i was able to determine in conversing with this person there is no way to disable http 3 and quick and chrome to disable it entirely there's some flags and chrome flags where you can turn some things off but i think you can't turn it off entirely so it's hard to do an apples to apples comparison but you pin your hopes on the smallest thing
John:
If Apple does plan on rolling out HTTP three or quick, their browser is going to break like Chrome is.
John:
And hopefully they'll be like, oh, geez, this browser has become useless because half of my HTTP connections fail.
John:
Let's see if we can fix this.
John:
As opposed to now where apparently no one is doing anything.
John:
Next possible fix.
John:
OSs and networking apps will need to be more careful about network changes.
John:
Apple will probably fix this in a Sonoma Point release.
John:
Other platforms I've seen this on are working on fixes.
John:
I wish I had this person's optimism.
John:
Will it be fixed in a Sonoma Point release?
John:
Does Apple even know about this?
John:
I haven't bothered reporting this because they're just going to say it's a Chrome error.
John:
And I'm not sure they're wrong.
John:
Like, I honestly don't know how much of the blame to apportion to Chrome versus Mac on this.
John:
Anecdotal evidence is that this is way worse in Sonoma.
John:
There are a bunch of new network-related demons for various things that people have assigned blame to.
John:
Oh, it's because of the U-ton thing.
John:
It's because of talking to a developer iPhone device.
John:
You know, it's because of Docker.
John:
It's because of a VPN.
John:
You know, it's because whenever your iPhone and your Mac are on the same Wi-Fi network, it happens.
John:
Like, pfft.
John:
I don't know, but I don't I don't know how to to communicate this to anybody involved other than to say Chrome no longer works on Mac in a reliable way, like in a fundamental, no longer useful as a browser kind of way, because, you know, some large percentage of your requests just fail and it's not great.
John:
uh finally the last thing there is an experimentation going on in cronet to improve things by migrating connections across network interfaces this is the migrate sessions on network change v2 flag or feature whatever we'll put a link to the uh thing in the show notes this isn't just a bug fix though it has trade-offs so it likely needs to be done in combination with improvements from os's and networking apps so the idea behind this thing is hey if there's some connection in this cronet networking library and the network changes and we have to
John:
you know not use that connection anymore because it won't work whatever was using that connection why don't we smoothly migrate it over to the new working connection and then the application that is using chronet which would be the chromium browser engine doesn't need to worry that that happened it will just continue to work that would be great but it's not just like it's not just a bug fix because there are performance implications of that migrating it is not free it takes time um
John:
And it's not the same as just not having the network change.
John:
So I don't know what the solution is here, but it's extremely frustrating.
John:
We did not have an error network change shirt as part of this sale, but I'm seriously thinking about it for future ones.
John:
This continues to happen.
John:
And then the final bet, this is really the icing on the cake.
John:
This is not seemingly related to error network change, but kind of is.
John:
This was a toot by someone named Thomas who said, I just learned what the user interface in the SpaceX capsules run.
John:
The capsules that provide life support for people traveling into space have to be absolutely reliable.
John:
The user interface that controls an explosion.
John:
It runs a home-compiled version of Chromium, and the UI is written in JavaScript.
John:
Now, this person wrote this to say, can you believe they're using web technologies and a thing that has to be reliable?
John:
But that's not the thing that sticks out.
John:
The thing that sticks out is the word Chromium.
John:
Do you think if they have Air Network changed in a SpaceX rocket...
John:
That might be a problem.
John:
And who's responsible for fixing that book?
John:
I'm sure it's running Linux or whatever.
John:
But again, people running Chrome on Linux or Chromium based things on Linux are also experiencing this problem.
John:
sometimes i feel like i'm the only person with this problem but every time i talk about on the show people send me toots on mastodon here i am i'm getting it i'm getting a network changed i got people on linux people on max so far no one in ios because maybe those errors are hidden from you on most ios apps but i mean i still think my window dragging bug is more important because i could just use safari but uh for the people uh in the spacex capsules that have a ui that's running chromium maybe talk to your bosses about this
Marco:
We are brought to you this episode by Squarespace, the all-in-one website platform for entrepreneurs to stand out and succeed online.
Marco:
Whether you're just starting out or managing a growing brand, Squarespace makes it easy to create a beautiful website, engage with your audience, and sell anything from your products to your content to your time all in one place and all on your terms.
Marco:
squarespace has an amazing platform to build websites it has never been easier for anyone to unlock unbreakable creativity with their awesome all new fluid engine you start with a best-in-class website template and then you customize every design detail you want with reimagine drag and drop tech that works on desktop and mobile you can really stretch your imagination online with fluid engine built in with all new squarespace sites
Marco:
And Squarespace has amazing online storefront support.
Marco:
So you can sell your products on an online store and not just physical goods, even digital goods or service products, or even things like time slots.
Marco:
Squarespace has all the tools you need to start selling online, all with flexible payment options, seamless checkouts that really convert very well, all sorts of options for people to pay, including, of course, credit cards, but also PayPal, Apple Pay, Buy Now, Pay Later, all sorts of great support there.
Marco:
All of this is backed by powerful analytics so you can grow your business with insights about things like where your site visits and sales are coming from, which marketing channels are most effective.
Marco:
You can improve your website and build a marketing strategy based on your top keywords or most popular products and content or whatever you want.
Marco:
You can try everything on Squarespace at squarespace.com and start a free trial there.
Marco:
You can see it all in trial mode and you can see how well it's going to work for you.
Marco:
When you're ready to launch,
Marco:
Go to squarespace.com slash ATP for 10% off your first purchase of a website or domain.
Marco:
So once again, squarespace.com to start that free trial.
Marco:
When you're ready to buy, squarespace.com slash ATP for 10% off.
Marco:
Thank you so much to Squarespace for sponsoring our show.
Casey:
As we record this, it is Monday the 20th.
Casey:
In the last three days, there have been approximately 7,304 different CEOs for OpenAI.
Casey:
Somehow still only one board.
Casey:
Well, we've been recording for a while.
Casey:
I don't really have too much to say about this.
John:
Just read the headlines in the notes.
Casey:
All right, so these are the headlines, and we'll link these in the show notes.
John:
I think this is like 48 hours span, 24, 48.
Marco:
Yeah, it's been almost no time.
Casey:
So these are the headlines as per The Verge.
Casey:
Sam Altman fired a CEO of OpenAI.
Casey:
That was on the 17th.
Marco:
And by the way, out of nowhere.
Casey:
Yeah, so let's get through the headlines.
Casey:
Yeah, I think it's worth unpacking all this.
Casey:
Then next, Twitch co-founder Emmett Scheer is the new CEO of OpenAI.
Casey:
That was apparently today.
Casey:
Microsoft hires former OpenAI CEO Sam Altman also today.
Casey:
Hundreds of OpenAI employees starting to resign and join Microsoft today.
Casey:
Is Sam Altman joining Microsoft?
Casey:
It doesn't seem to know.
John:
I think there's hours between those.
John:
How can you have it?
John:
Yes, it's a comedy of errors.
John:
But then every time there's a headline, it's like, this is a thing that happened.
John:
And it's like, but has it?
John:
Yeah, exactly.
John:
I skip the ones where it's like, oh, he was fired, but there are negotiations for him to come back.
John:
But no, he's not coming back.
John:
But yes, he is coming back.
John:
But they're talking about coming back.
John:
But now he's at Microsoft.
John:
But has he been hired by Microsoft?
John:
And this is not just like, oh, we're wondering what's going on.
John:
That's why the headline says, is Sam Altman joining Microsoft?
John:
Satya Nadella, who, by the way, is the CEO of Microsoft, doesn't seem to know.
John:
Satya Nadella announced, hey, Sam Altman's coming to Microsoft.
John:
But then when asked about it, like, hours later, it was like, well, not really sure.
John:
I don't check back in a few hours.
Casey:
Yeah, so let me try to give the quick executive summary of this.
Casey:
Now, I'm surely going to get a little bit wrong, but the general gist of it is, I think it was late Friday, Eastern time, there was a blog post from the OpenAI board, basically, that said that Sam Altman, who was the former CEO, maybe by the time you're listening to this, might be the CEO again.
Casey:
Who knows?
Casey:
But the former CEO had basically not... What was the phrase they used?
Casey:
Not been...
Casey:
forthcoming wasn't the word they used, but it was something along those lines.
Casey:
It was pretty severe.
John:
It was vague, but normally when a CEO is fired, this is the privilege of being in the executive ranks.
John:
When a regular person is fired, you just get fired.
Right?
John:
But if you're a CEO, no matter how terrible whatever you did was, no matter how badly you screwed up, you're always like, you know, moving on to spend more time with your family or want to pursue other interests or like they have some euphemism about how you're – it's not like we're firing him.
John:
It's a mutual decision we've all come to and he's decided to leave and he's, you know, pursuing his passions in this other realm and we thank him for all his blah, blah, blah.
John:
And this wasn't like that.
John:
This is as close as you're ever going to get to saying we fired him because we don't like him.
Casey:
Yeah, it was.
Casey:
That's how I remember being with forestalls.
Casey:
Basically, like it was clear that he got fired, but everyone put on the happy face.
Casey:
We would like Scott to spend more time with his family.
Casey:
yes, I would like to spend more time with my family.
John:
And usually, in those type of statements, there's something from the person who was fired, like, quote, you know, whatever, quotes from them, right?
John:
They'll say, like, I'm happy that I spent all my time with this company.
John:
I'm glad to be moving on, blah, blah, blah.
John:
Like, it's a participatory process.
John:
But what happened with Sam Altman,
John:
uh the story i heard on a podcast i was listening to is that he was on stage doing something like in his official capacity as ceo and he said okay well you know thanks everybody i have to go i have a meeting in that meeting he he got on he got on a call this is i was listening to the verges podcast about this he got into the call using google meet and one of the people on the show said even for 10 billion dollars no one will use teams uh because he's
John:
Microsoft invested $10 billion or whatever in OpenAI.
John:
So anyway, he goes to his meeting.
John:
He sits down and it's an online meeting, right?
John:
And all they did in the online meeting was read him that blog post you just referred to, Casey.
John:
They just put in that.
John:
That's what you do in these meetings.
John:
You have a script.
John:
You stick to it so you don't get sued.
John:
You just read the words.
John:
So if you're wondering, like, oh, they must have had an extensive meeting in which they explained to him why he was fired.
John:
When we say this came out of nowhere, it didn't come out of nowhere to us on the outside, because what do we know?
John:
It came out of nowhere to Sam Altman, who just got done doing a CEO thing and said, oh, I've got to go to a meeting, and they just read him a thing and said, yeah, you're fired.
Marco:
And by the way, the thing he was doing, having their whole developer conference, was widely lauded by the industry as an amazing thing.
Marco:
Everyone thought he did an amazing job.
Marco:
This is a great path to be on.
Marco:
Everyone's excited about OpenAI.
Marco:
Everyone's super excited about OpenAI and him.
Marco:
That's why...
Marco:
When this news dropped that the board fired him, everyone in the entire tech business was like, what?
Marco:
Are you serious?
Marco:
Is this a joke?
Marco:
Like, what happened?
Casey:
Yeah, so here's from the blog post I was able to dig it up while you guys were talking.
Casey:
This is the blog post from the OpenAI board.
Casey:
Mr. Altman's departure follows a deliberate review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities.
Casey:
The board no longer has confidence in his ability to continue leading OpenAI.
Marco:
Which, by the way, that's a really serious accusation.
Marco:
Potentially misleading the board is, in many cases, a crime.
Marco:
That is a really serious accusation.
Marco:
That's why this isn't just, oh, we had creative differences, we don't like them anymore.
Marco:
That's a very serious thing to throw around, what they threw around.
John:
And by the way, as far as I understand it,
John:
The board doesn't have to have a reason.
John:
Like, they have complete control.
John:
There's no, like, power struggle or whatever.
John:
The structure of this company is such that the board can decide, oh, we don't want you to be CEO.
John:
I think for pretty much any reason.
John:
Like, he needs to do anything wrong.
John:
It happens all the time.
John:
The board's like, we want a different CEO.
John:
We want to go in a different direction.
John:
That happens all the time.
John:
They don't need to explain, which is why you can do the, oh, they're going to spend more time with their family or whatever.
John:
Whatever the real reason is, it's not like they need to make something up.
John:
Like, we can't fire him unless we say that he lied to us.
John:
They could have just said later, but they threw that in there to say, oh, and we have reasons.
John:
We think he has not been, quote unquote, consistently candid.
John:
Like they didn't have to put that in there for any reason other than to just I mean, I don't know.
John:
Maybe this will give us cover because it's vague enough.
John:
People can imagine some terrible thing happened.
John:
But bottom line is apparently a majority.
John:
We don't even know how much because it's not it's not a public company.
John:
So it's not like they have public records to say, was this a unanimous vote?
John:
Did everybody vote to kick him out or whatever?
John:
But a majority of the people on the board decided they don't want him to be CEO anymore.
John:
And they made this decision without consulting Sam Altman at all.
John:
I mean, maybe they've had press conversations about it.
John:
Maybe he should have known it's coming because they said in their last meeting, if you if you do really well in that dev conference, you're fired.
John:
I don't know what they might have said to him.
Casey:
So the impression I get in – I don't have a lot of – nobody has facts here at the moment.
Casey:
But the impression I got based on the coverage that I've read is that there were two different factions or tribes, I think Sam had called them at one point or another, two different tribes within OpenAI.
Casey:
OpenAI was originally founded in 2014, 2015, thereabouts, as a nonprofit.
Casey:
And their theory was we want to make, you know, basically artificial general intelligence.
Casey:
We want to make, you know, AI.
Casey:
But we understand that with this great power comes great responsibility.
Casey:
And we want to do this to improve the lives of people everywhere.
Casey:
But we're going to try to do it very methodically, very deliberately, very safely.
Casey:
And that was how the company was founded.
Casey:
And it's not even really a company, I guess.
Casey:
It's a nonprofit.
Casey:
It's a nonprofit, right?
Casey:
So that's how it was founded.
Casey:
Well, at some point, you know, Altman swoops in.
Casey:
I don't recall exactly when it was in the timeline, but it's clear that Sam Altman is very...
Casey:
Silicon Valley VC, which I personally find to be a very ugly stereotype.
Casey:
I just don't care for that whole grow, grow, grow, take over the world mindset.
John:
And he is the embodiment of that stereotype.
John:
It used to be with Y Combinator, which is also kind of a poster child for that type of VC.
Marco:
Yeah.
Marco:
And he's not the guy who had the day of phone, night phone.
Marco:
He's the guy who wore the multiple polo shirts at the same time.
Marco:
Yeah.
Casey:
Exactly.
Casey:
Now, I'm trying to make plain, to quote Merlin, my priors on this.
Casey:
I find that whole Silicon Valley mindset to be very off-putting.
Casey:
you know, grow, grow no matter what.
Casey:
We don't care who we run over in the process.
Casey:
Grow, grow, grow.
Casey:
You know, let's make money, make money.
Casey:
And that's all that matters is making money.
Casey:
Nothing else matters.
Casey:
Grow and make money, grow and make money.
Casey:
Or sometimes you don't even have to make money.
Casey:
You just grow.
Casey:
Well, so anyway, so consider that when I, when I give one of my, you know, lip turns up as I tell you this story, but nevertheless, so Sam comes in and he's grow, grow, grow, money, money, money.
Casey:
And
Casey:
everything i guess everyone was kind of coexisting okay and even though they disagreed internally everything was mostly all right well then they released chat gpt what about a year ago and it's it it's it makes a tremendous splash it's i think a lot of people are saying it's the quickest adopted consumer product ever because you know they had something like 100 million signups in the span of like four minutes not literally but you know what i mean
Casey:
And so now money is becoming a thing, like for real.
Casey:
And suddenly we have to pay the piper on this division between money, money, money, and let's do this for the good of all people.
Casey:
And their chief scientist, Ilya something or other, I don't have the name in front of me, I guess was more on the let's play it safe, let's be deliberate, let's do this for the good of humanity.
Casey:
And Sam was very much money, money, money.
Casey:
And at some point...
Casey:
it appears something gave.
Casey:
And the rumblings that I heard, not from sources or anything, just based on the reporting I read, was that Ilya had convinced the board, of which I think he is a part of the board, let's give her to Sam.
Casey:
And then all hell broke loose.
Casey:
And now as we record on Monday night, we don't really know what the latest and greatest is.
Casey:
But I understand, you know, both sides of this.
Casey:
Like on the one side, if the board really did establish or if OpenAI really was established as a nonprofit, which I think is factual, I don't think that's up for grabs.
Casey:
It is within reason for them to say, well, hold on.
Casey:
Suddenly we've taken a turn and pivoted to money, money, money.
Casey:
We don't like that.
Casey:
And Sam seems to, by most metrics, have pivoted them in that direction.
Casey:
So if we don't like this pivot to money, money, money, then we probably don't like Sam anymore.
Casey:
So on the surface, I don't have a problem with that.
Casey:
And again, my priors tell me, yeah, screw that Silicon Valley nut job.
Casey:
Let's do this for the good of people rather than the good of the almighty dollar.
John:
Well, wait until you hear about the other side of that, because I think they're both not jobs.
Casey:
Well, that's fair, too.
Casey:
That's fair, too.
Casey:
But I'm just trying to make plain.
Casey:
This is my biases coming to light.
Casey:
That being said, unquestionably, ChatGPT and the work of OpenAI and DALI and all this AI stuff, whether or not you think it's cool, whether or not you think it's good –
Casey:
I think unquestionably it's important and it's been making a big damn splash.
Casey:
And I think that there's a lot of interesting things here.
Casey:
And unlike blockchain, I think there's a lot of fascinating threads that we can pull.
Casey:
And I think that there's a lot there.
Casey:
And because of that...
Casey:
I also have sympathy for the, let's grow this product and see what the world can do with it.
Casey:
Let's not slow down.
Casey:
Let's not be deliberate about it.
Casey:
Screw it.
Casey:
Let's just figure it out as we go, which is when Silicon Valley is at its best.
Casey:
So I have very mixed feelings about this.
Casey:
And I mean, as someone who likes drama more than I should, because I'm a
Casey:
I really shouldn't enjoy drama this much, but oh, this is delicious drama.
Casey:
And I am here for the drama of it.
Casey:
But I honestly don't know who's right, who's wrong.
Casey:
I'm not sure any of us do.
Casey:
I don't know what to make of this, but it is a mess.
Casey:
And it's been a mess every other hour since late Friday evening.
John:
Yeah.
John:
By the time you hear this episode, who knows what will have happened.
John:
I think this is actually, I mean, the most Silicon Valley thing about this
John:
is that a lot of the companies that you know from my youth that sort of came out of silicon valley they were at the forefront of some technology that they were the right place at the right time right they decided to make personal computers and personal computers were just becoming possible right you know you got your apples your intel right you know microchips memory you know cpus x86 microsoft with the software like they're on they're riding a wave of something
John:
And those particular companies, especially, you know, we know a lot of the names of the founders, Bill Gates, Steve Jobs, a lot of them were younger people.
John:
And pretty much none of them went into it thinking, I'm going to found what's going to become what we know today as tech giants.
John:
Because that's not what, you know, if you tried to tell a young Steve Jobs and Steve Wozniak what Apple will look like today, they'd be like, yeah, right.
John:
No one thinks that's going to happen.
John:
So what happens with these companies that are incredibly successful is they're made up of people, the founders, but also usually they bring in other people to help them run the company.
John:
And especially for companies that experience explosive growth, the people who are responsible for running that company, they're just people.
John:
Sometimes they're people who've never done anything like this before.
John:
And so if you're wondering, how can this company that's been so financially successful and apparently has such amazing technology not have their governance straightened out and basically conduct their business in this ridiculous way with this whipsawing of the CEO being fired and asking him to come back and regretting telling him to leave and all that other stuff, it's because there are young people
John:
tech company riding the wave of new technology run by a pinch of people who are bad at doing this and that's i mean look steve jobs was fired from apple by a bunch of people that were brought on to run the company that was probably the wrong decision for them to do at that time they if they had you know nurtured his town perhaps they would have had a more successful 90s but you can also understand why they fired him because you know read the history like a lot of these companies in their early days at least run up to the edge of
John:
being you know doing something terrible but being run in a way that is not i'm not gonna say not sane but that is that is not what you the way you expect a very wealthy company to run that's even with google they have the two founders and we have we have to bring in the adults to help run things and you know it's you can go it can go badly in so many different ways it's very rare that you get a company
John:
that manages to get through this sort of awkward adolescence and sustain its success into something that continues and becomes like a company that is more reliable and steady and does not have weird boardroom drama like this.
John:
So I'm not saying that this is unique to OpenAI at all.
John:
Like I said, I think this is the most Silicon Valley thing that they've done.
John:
That said, the people in these two camps here, the Sam Altman and the OpenAI people,
John:
are a little bit weird.
John:
Both sides are a little bit weird.
John:
So we didn't talk about the Microsoft stuff, but like OpenAI is this nonprofit.
John:
Microsoft wanted to do a thing with them.
John:
So here's the thing with OpenAI.
John:
Their goal is to make artificial general intelligence, which is like pound 9000 or whatever, but it doesn't kill you.
John:
That's a big asterisk there.
John:
Yeah.
John:
They haven't done that, to be clear.
John:
That's like their aspirational goal.
John:
It's like their mission statement.
John:
But what they did do was make ChatGPT.
John:
And it turns out that ChatGPT is a useful thing that people can do stuff with.
John:
and people everyone's interested in it and microsoft was interested in it and they said hey we would like to do stuff with your technology too the problem open ai has is they want to make artificial general intelligence but turns out doing anything even approaching that costs tons and tons of money and so open ai's idea was like we'll just we'll raise money somehow and people will give us money because they know we're doing the good work to make hell 9000 not kill us and
John:
And the amount of money they were able to raise was a fraction of the amount that they would need.
John:
Microsoft came and said, we kind of like that things you're doing over there.
John:
How about we let you use our massive computing resources and our data centers?
John:
to the tune of billions of dollars worth of free credits, like little token, Chuck E. Cheese tokens, now you can run your stuff in Azure.
John:
Because OpenAI can't do anything without large amounts of computers, which takes large amounts of money.
John:
And they're a nonprofit, and their idea was like, we'll raise that money.
John:
People will give us money to pursue this.
John:
They just did not get as much money as they would need.
John:
But ChatGPT, Microsoft's like, hmm, kind of like that.
John:
So they did this deal where Microsoft is like...
John:
Giving them $10 billion, most of which is in the form of credits to run stuff in their Azure, you know, cloud computing stuff.
John:
And but how can they how can they do that?
John:
Well, if you look at the org chart, it's like there's a nonprofit and the nonprofit controls this other for profit thing that gets money from Microsoft.
John:
But the for profit part, it's not it's not I think it's like what profit capped or something like that.
John:
Anyway, it's entirely controlled by the nonprofit.
John:
But still, Microsoft has this financial interest.
John:
And by the way, as part of this deal, Microsoft gets all the rights to the the open AI IP, like their intellectual property, whatever.
John:
I don't know what their intellectual property is, but presumably whatever it is they use to make chat GPT.
John:
Microsoft now has the rights to that.
John:
i think forever as part of this 10 billion dollar deal the only thing microsoft doesn't have the right so this is another one of those great deals kind of like uh microsoft uh doing the internet explorer thing and saying oh don't worry we'll give you x percent of all of our internet explorer sales and then giving it an internet explorer away for free to everybody and so that person got nothing anyway um the company that got nothing from that
John:
Open AI is like, okay, we'll do this $10 billion deal.
John:
We'll get access to your computing resources, which we kind of need to like literally do anything because we don't have enough money to do this AI stuff.
John:
We don't have enough money to pay for the computers to do it.
John:
But what we won't give you, and we'll license you our current technology, but what we won't give you is anything having to do with artificial general intelligence.
John:
So if we invent HAL 9000, Microsoft, you don't get it.
John:
And Microsoft's like, okay, I guess.
John:
And Microsoft's actually thinking, they're never going to do that.
John:
It doesn't matter.
Right?
John:
So they retained the right to the fantasy thing that they want to make, but gave Microsoft the rights to all the other stuff.
John:
And so Microsoft's doing all this stuff with their co-pilot, ChatGPT, all that stuff, ChatGPT, the fact that it does useful things, that is a product that can make money.
John:
Microsoft can incorporate that technology into its own products, Microsoft Office, GitHub, everything, and use that to make money.
John:
That's what everyone's doing.
John:
Meanwhile, the nonprofits over there are going,
John:
What are they doing making products and making money for?
John:
Don't they understand we're trying to make Hell 9000 and we're trying to make sure Hell 9000 doesn't kill us?
John:
And so there's already a disconnect.
John:
And so on the one side, it is like you have a technology that people will pay money to use.
John:
I forget what their numbers are, but like they got 100 million users.
John:
They're making some huge amount of revenue to it because they charge for access to this stuff.
John:
And of course, Microsoft, to the extent that Microsoft incorporates any of this technology into their products, that helps them sell more of their products, which of course they make money on.
John:
So it's, you know, it's a product business.
John:
hey we have a thing we came up with where it was google search or you know whatever and we can use it to make money and that's what sam altman's out there doing using a product to make money and then the open ai people who are like ai is going to kill us we need to create it in a first we need to create it but we need to create it the right way so it doesn't kill us and everyone else is like yeah but you haven't created it but we've got this thing over here called chat tpt that people want to pay for that's not ai even though everyone calls it that
John:
But it is a product that people want to use.
John:
So can we have a developer day and make an API and charge people for API access and do deals with Microsoft?
John:
And it's like, OK, I guess.
John:
But eventually, like, no, we're making L9000.
John:
We don't like that other stuff.
John:
Stop it.
John:
And so now you have this disconnect.
John:
And if you think about what OpenAI is, it's this mission statement.
John:
It's a bunch of employees that implement this mission statement.
John:
And then it is the output of the knowledge and the output of those employees.
John:
And so Sam Altman, apparently, was popular within the company.
John:
So when he left and went to Microsoft or did he of the 700 or so employees at OpenAI, about 500 signed a letter that said, hey.
John:
If you don't bring him back or if you don't all quit or whatever, we're all going to go to Microsoft because Microsoft said we have an open standing offer to go work for them instead.
John:
So if Microsoft gets Sam Altman and 500 of the 700 employees, what is the nonprofit left with?
John:
Microsoft already has the rights to all the IP of everything they actually made because they didn't make HAL 9000.
John:
So they've got the rights to that.
John:
And if all their employees also go over there and they have the former CEO, an open AI has a mission statement and 200 loyal employees left.
John:
And really, I guess those people can regroup and use their $10 billion of Azure bucks to figure out how to make HAL 9000 any day now.
John:
Meanwhile, Sam Altman and Microsoft are over there continuing to sell access to ChatGPT to autocomplete stuff when you type reminders in or whatever the hell they're doing to make money.
John:
The reason I think both camps are a little bit weird is because Sam Altman's in the, as Casey alluded to before, the grow, grow, grow, boil the ocean, kill all the poor people so we can make another buck because in the end, what really matters are the people who are going to live a trillion years from now.
John:
lots of interesting philosophical exclusives to be a jerk today but really it's it's for the future anyway money money money and then the flip side the people will think it's going to come and kill us all and they need to be really careful about how they create it and you're like what what's going to kill us all like other thing no one's been able to invent yet but when they do it's going to be really bad it's like yeah uh dragons can kill us all too but uh there aren't any dragons like but there could be you're right there could be someone could genetically engineer a dragon but do you have a dragon no
John:
but i think i know how to make one all right well get back to me about the dragon thing yeah so it's kind of like self-driving cars it's like any day now it'll come it's like well no one's ever done it but all these companies sprung up about about the promise you know like based on the promise of like well you know in five years there'll be safe self-driving cars so we need to build business around that reality and that reality has not happened and so a lot of those companies are having problems so open ai is trying to pursue how 9000
John:
But I don't think they know how, like everyone else who has tried before them.
John:
But unlike lots of other companies, they've made a useful thing that is a marketable consumer product in the course of trying to do that research.
John:
And it seems that that useful consumer product is not compatible.
John:
With their mission statement, because that's not what they wanted to do.
John:
Oh, that's cool and all.
John:
That's a step along the way.
John:
But let's not get distracted by trying to make a multi-billion dollar business out of this useful thing that we've made.
John:
And by the way, OpenAI is not the only company that's done this.
John:
Lots of other companies have similar things.
John:
So it is kind of a cutthroat business.
John:
OpenAI seems to be the leader in this field, which is why Microsoft is interested in them and why they did that deal.
John:
But it's not like this is unknown elsewhere.
John:
What is unknown everywhere is HAL 9000.
John:
that doesn't exist don't and so if you're scared of it existing okay it would be scary if it did exist but it doesn't and i'll be more interested once you think you have a road to making it like can you go from chat gpt to hell 9000 maybe maybe this is step one of five maybe step one to 5 000 or maybe you're just barking up the wrong tree and this approach isn't going to work anymore the planes that flap their wings we'll see
John:
but uh yeah that's that's kind of the state of the world and these two camps at least the the uh the sam altman thing you're like okay i've got this guy's number i've seen people like this before i know what he's doing he's making products he's selling them it makes sense and the opening the eye of people are more like very confused sort of monks who have noble if misguided goals and great ambitions but what they increasingly don't have are people and money to accomplish that
John:
And it's all through their own kind of, I'm going to say mismanagement, but through their own apparent misreading of the situation, which is, what did you think you actually have?
John:
Oh, we control the company.
John:
We can fire him.
John:
But if all the employees log to him and he goes to Microsoft and all your employees leave,