whynotminot

What you personally believe and perception within the United States are not the same thing.

Doesn’t Broadcom bring a lot more to bear here than just Verilog? Including relationships with the actual fabricators.

I’m a little surprised to see how much trouble people in this thread have had with the Cal Digit TS3.

Mine works pretty well — have used it with three Intel MacBooks in the past and now currently two different Apple Silicon MacBooks.

One of the Intel MBPs did not like it. Would reboot every time I unplugged it from the dock. I blamed that MacBook for that one, since nothing else was ever a problem. I sent every crash report to Apple, along with some choice words that my $2,500 MacBook should be able to handle connecting to a very commonly owned TB Dock. Eventually they did fix it and it stopped being an issue.

Has ended up one of the more reliable pieces of tech gear in my life, especially given the absolute mad complexity of TB3 behind the scenes.

This model might trigger planned obsolescence legislation in some jurisdictions.

As opposed to years of garbage Windows laptops at the $600 price point, which you seem to expect will remain viable longer than the Neo ... ?

If any legislation comes out of this, it will simply be because of Apple's high profile. They were content to let e-rot fill the shelves for years before the Neo.

What does “main” mean? Main processor for a flagship iPhone? Or main processor for a HomePod?

There’s a lot of “main” processors for Apple’s devices at this point.

I would be deeply skeptical of a brand new flagship iPhone <n> Pro having an Intel fab’d SoC until at least a few years into this arrangement.

Big deal, smart for all parties, really. Apple standards will make Intel step up and become a better foundry partner.

Apple will gain increasingly needed diversification.

US supply chain gets a boost.

Should be fine for TSMC in the short to medium term. Apple not going to risk actual mainline iPhone SoC on Intel any time soon, so lion share of TSMC Apple revenue will be fine.

We seem perfectly fine splitting up some aspects of adulthood, like 21 for drinking.

Sure, raise it past the transition period.

I’m tired of the pearl clutchers. Decide an age you’ll actually accept. That’s an adult. No more infantilization.

Bought and read the book and honestly the author is just as crazy as the rest of the executive team she skewers. I’m not too broken up that she didn’t get more money from my purchase.

Can we raise the age of adulthood from 18 to whatever acceptable age ends this discourse once and for all?

You’re describing all the advantages that Russia in theory had when it invaded Ukraine. That war remains in stalemate.

With US support and the resilience and ingenuity of their people, Ukraine has persisted.

It is possible to land large numbers of boots on the ground.

I think you need to do more research on how challenging a Taiwan invasion would be. It is nowhere near as simple as “just cross the strait. Put those boots on the ground.”

There is a reason it has not happened. It would be incredibly logistically challenging.

Sure you can take an easy/undisciplined target like Maduro.

You think what the US military did there was easy? lol. Lmao, even.

I agree with your statement that human resilience can outlast a better army.

But then you go on to say:

Which is why if China makes a move on Taiwan, the US can do nothing.

If your opening thesis is true, then it's strange you follow it up with that. Taiwan has just as much a chance of outlasting a stronger competitor as those other countries that resisted US dominance.

And with the US providing them weapons, intelligence, and support, maybe a better chance. See Ukraine.

I think that says more about our political leaders than our military.

Politicians choose the war and our military fights the battles. We're very good at winning battles. But some wars can't be won. The problem then lies in their choosing.

I imagine Sisyphus became the best, most effective rock push in the world. Unfortunately despite his talents, the task he was assigned was insoluble.

Pretty sure anyone who fights the US military finds out pretty fast it’s a good military.

It isn’t perfect. It has flaws. War is hard to get right in every dimension.

They bought them for the streaming service that came with it. Not for fake weights in headphones.

Love this kind of behind the scenes -- thanks for sharing!

Personally, I think in a way you both ended up right. Content is king. But with the iPad (my favorite PDF viewer) as an important part of the Preview landscape now, I view the right side as where content should live.

I still love Opus but it's just too expensive / eats usage limits.

I've found that 5.3-Codex is mostly Opus quality but cheaper for daily use.

Curious to see if 5.4 will be worth somewhat higher costs, or if I'll stick to 5.3-Codex for the same reasons.

As long as both companies remain stable and viable, there's probably limited upside to pouring more money into them. If they fail, and bring down the AI ecosystem with them, that is very bad news for Nvidia. So they've been there nurturing their success and providing capital to backstop their exponential growth.

You can see Nvidia stepping in throughout the ecosystem with confidence boosting investments where needed. They haven't just supported Anthropic and OpenAI.

If OpenAI and Anthropic succeed, and get their business fly-wheels fully spinning, they don't necessarily need more capital from Huang. Ultimately the goal of Nvidia is to profit from their long-term success by selling them GPUs for a long, long time. The goal isn't to keep plowing money into them forever.

You could get something smaller but have it closer to your face than 1m?

The sort of “visual impact” a screen can have is mostly a combination of what percentage of your FOV it consumes.

People think they’ve got a bunch of screen real estate when they buy a big TV to use as a monitor… and then they use it a twice or more the distance of a regular monitor.

I get it. I’m scared too. I’d be lying if I said I wasn’t.

Actually it is not about this stage. It is about the sustainability of this when training data runs out

This is an argument from 2024. Somehow, the models have continued to improve.

If they stopped improving today they are good enough as they already are to generate profound change.

The wave front is already visible, we’re just on the shore waiting for the impact.

Personally at this point my combined AI spend is the most expensive recurring monthly subscription I have, and that’s even with my company also paying for the AI tools I use at work.

If it weren’t subsidized I would pay more. Wouldn’t be happy about it but I would do it.

At this stage in the game I don’t really understand where this skepticism of the value these tools provides comes from.

What? SoftBank has been investing in them repeatedly for years now.

This lazy kind of post annoys me because it sort of groups any of us saying that this technology is profoundly different in with all the town criers who have said this kind of thing before — even if we have never said it before and were even skeptical of past declarations

Effectively, it’s a statement saying nothing can ever be profoundly different, because people have said it before and been wrong.

Lazy.

I would guess there's not enough volume due to limited use-cases of the tech compared to more traditional screens.

The typical e-ink uses cases boil down to e-readers, dumb-phones, and hobbyists, which is not a huge market. Anything niche or specialized tends to carry a higher cost.

In my org, oncall is just who gets the 2am phone call. They can *try* to contact anyone else if needed.

Emphasis mine.

Claude is there as long as you're paying

If you’re at a company that doesn’t pay for AI in the year 2026, you should find a new company.

and I hope he doesn't hallucinate an answer.

Unlike human coworkers with a 100% success rate, naturally.

Do you know what on call means?

It means Sam is 7 beers deep on Saturday night since you’re the one on call. He’s not responding to your slack messages.

Claude actually is there though, so that’s kind of nice.

Yeah. Everyone sort of assumes that not having personally written the code means they can’t debug it.

When is the last time you had an on call blow up that was actually your code?

Not that I’m some savant of code writing — but for me, pretty much never. It’s always something I’ve never touched that blows up on my Saturday night when I’m on call. Turns out it doesn’t really change much if it’s Sam who wrote it … or Claude.

The hallucination rate with the Gemini family has always been my problem with them. Over the last year they’ve made a lot of progress catching the Gemini models up to/near the frontier in general capability and intelligence, but they still felt very late 2024 in terms of hallucination rate.

Which made the Gemini models untrustworthy for anything remotely serious, at least in my eyes. If they’ve fixed this or at least significantly improved, that would be a big deal.