Things weren't simpler. The complexity was simply not visible because different teams/department were all doing a small part of what now a single team is doing with Kubernetes. Yes, for that single team it is more complex. But now it's 1 team that does it all, instead of 5 separate teams responsible for development, storage, networking, disaster recovery, etc.

Kubernetes is a gift.

Exactly. The GPT 5 answer is _way_ better than the GPT 5.1 answer in the example. Less AI slop, more information density please.

I find it strange AGI is the goal. The label AI is off and irrelevant. A language model is not AI, even a large language model. But language models are still extremely useful and potentially revolutionary. Labelling language models as AI is both under and overstating the value. It's not AI (insert sad trombone), but that doesn't mean it's amazing technology (insert thunderous applause).

Without defining AGI as a goal, current AI companies would not be able to amass the amount of money they want

This terminology is confusing. Historically, AI was always used to mean any kind of machine intelligence, including the most basic novice chess AI, or an image classifier, or a video game character's AI. Now a lot of people seem to be using it as a synonym for AGI - a human-level intelligence.

Greg Brockman talks about it on Lex’ podcast I think. He thinks ultimately AGI won’t be anything more than token generation, if I remember correctly

Scalability is not purely technical. It's also organizational. For all its drawbacks, the microservices architecture is easier to scale from an organizational perspective.

Only when the service boundaries and interfaces are built with this in mind.

A service that is isolated enough it could be another company? Sure, this scales. But do company hierarchies and organization practices help this happen? I haven't seen it outside of places like Amazon where there was a mandate for it to be that way.

What companies end up with in practice are services so tightly coupled with the rest of the company that they requiring a mishmash of API requests in both directions and endless coordination. Aka a distributed monolith. All the problems with zero the advantages.

My experience is it adds a multiplier 10x to the work that needs to be done, which is then a good fit for the 9x extra developers upper management decided to hire.

Sure, it keeps everyone busy with work they feel is meaningful -- but it really is work that was not really needed in the first place with a monolith.

Much simpler to try to stay fewer developers as long as possible.

Also in my experience it is impossible to get people along with organizing the company along software architecture division lines, thus creating a lot of political problems.

This is a strange perspective. Indeed, no one is paying for Copilot, but no one is paying for AI in general. No one is making any profit on AI. This is not a Copilot problem, but a characteristic of the AI market in general. Only the ones who are selling shovels are making money (nVidia).

It has already been trained on all the data. The other obvious next step is to increase context window, but that's apparently very hard/costly.

I don’t think this is true. See https://arxiv.org/html/2211.04325v2 for example.

There are diminishing returns.

What do you mean "future"? This is already the case. M-series chips share the same architecture with A-series CPUs. If you increase the A19 core count you'll end up with something close to M5.

The ThinkPad E14 he mentioned is available with a 120Hz 2880x1800 display.

I think it wasn’t out 5 months ago when I looked. But that honestly looks like a great option now

It's everything else where MacBooks excel too. The build quality is insane, I've never seen a laptop with as little flex as a MacBook Air or MacBook Pro. The keyboards are now finally great, and the display is amazing as well. Trackpad is best in class too.

Another thing: the displays are glossy, but still not very reflective. The Windows laptops glossy displays are so much more reflective, they are unusable outside. Also something worth mentioning, the MacBook displays get really bright. A high-end OLED display hardly goes above 400 nits. A MacBook Pro can go to 600 nits and outside it goes to 1600 nits. This is the difference between being able to use a laptop outside, and not.

Durability: if you're not doing anything crazy the MacBook will look brand new even after years of usage. Notable exception is the cheap plastic key caps which degrade very quickly, a bummer.

So the MacBooks beat the competition easily from a hardware quality perspective, and we haven't even talked about the elephant in the room yet: CPU performance, battery life and fan noise, obviously Apple is even further ahead in this area.

And then price, as strange as it sounds, both the MacBook Air and MacBook Pro are cheaper than the competition.

Disclaimer: I don't own a MacBook personally because I think macOS is not great, but that's probably the only reason why I'm not buying a MacBook. I would happily pay same the price of a MacBook Pro for a similar Windows laptop if it existed. It does not. There are always compromises.

There is no obligation to contribute back. That's the whole point of open source. It's irrelevant how much WP Engine contributes.

Also, uh, they did contribute back, and to the ecosystem.

That they didn’t is pure lies and FUD from Matt who retreats back to the position that it “Wasn’t enough”…