Great point. It's also really unclear how many Mythos found that are not/would not be found by open weight models.

This is pure marketing FUD, which is one of Anthropic's most advanced forms of "intelligence".

While I agree that many in the industry want the cheapest and the fastest to build regardless of quality, my question is about the demand side.

Who wants these apps? What is is that they do that someone is willing to pay for? How does that address the cost of the other inputs that make apps worth enough money or other rewards that someone wants to maintain them?

We are 18 years out from the launch of iPhone App store, and even though humans are far slower than AI in building apps, after nearly two decades I don't think there are massive parts of human activity that are un-apped. In pharmaceutical development, the dearth of "undrugged diseases" has led pharma companies to focus on rare and orphan diseases - bringing VERY high cost drugs to market to serve small numbers of people.

Where are the "orphan applications" that these apps are there to serve?

Vibecoding a delivery app stack will not make DoorDash obsolete - somebody still has to recruit drivers and food sellers and offer something to each of those parties that makes them want to drop DoorDash. DoorDash may be able to automate away some labor (though I suspect it will be less than they think).

In the enterprise, the theoretical "un-automated work" seems to be in two main buckets:
1- making presentations, dashboards, documents automatically, and
2 - building software automatically

Both of these clearly have some value, but I think the AI boffins and investors are wildly overlooking all of the human stuff that goes along with those tasks.

Also, it's obvious that AI makes that kind of stuff a commodity, meaning that its value goes down as its volume increases. So yeah, AI can make a lot of slop, but it's not obvious how that makes there be more valuable stuff.

Yeah, agreed. I actually am really struggling to understand how they think there is demand for this kind of thing. Is there market research where someone says "sure, I don't care who's talking, just as long as the content is a topic that I am plausibly interested in and the _conversation_ is not too jarring".
It's true that people will watch AI slop on YouTube; my guess is that is the demand signal they are responding to.

I have once watched an AI generated YouTube video (really it was weird graphics over an audio that sounded like Richard Feynman reading) and I found the explainer vaguely interesting enough to continue watching in the background for 20 min while I worked. I guess that channel got a few ad impressions off my eyeballs for that.

So I can see notionally how this kind of slop might win over some other slop in a transactional zero-sum way, but I really can't see how any of the typical marketing stuff - audience building, brand development, downstream sales conversions, fits in with AI-generated, undifferentiated commodity content. Isn't the whole idea of an influencer that the audience wants a person to connect them to content?

Maybe it will work, but my guess is it will just result in a lot more content that one has to wade through. I'm sure someone will have an AI solution to that. Turn that Hamster wheel up to ludicrous speed!

You are dead right. Agnosticism is much closer to a "null hypothesis" than atheism, yet many atheists like to present their view as "more scientific".

A really rigorous scientific examination pretty quickly comes to the conclusion that the existence of deities (in the spiritual, moral, and experiential ways that most deities are defined) is essentially impossible to prove or disprove with scientific examination of the material world.

In my own opinion it's because faith and religion are mostly about questions of meaning and purpose rather than physical assertions. One's reaction to the tenets of a religion are also experiential - if those tenets provide meaning clarity and improvement in one's being they make sense.

The other aspect of agnosticism is that I find attractive is a kind of intellectual humility. As a single, limited human being, who I am to say what is absolutely true in the metaphysical planes of existence?

I think his age is a big factor even without dementia.

Many older people I've seen have a much stronger "wow, look at this reaction" to new technology- they still see advances as minor miracles when myself and others see them as incremental progress.

I also think a lot of people in that generation have spent less time with psychology and are less prone to being aware of their own emotional and cognitive reactions. This means they don't spend much time reflecting on whether that "wow" reaction is a factual assessment or just an emotional reaction.

And of course the LLM vendors love having people debate this stuff, it keeps the "AI is magic" aura alive even when so much of what "AI" is making seems to be slop.

...and the Cybertruck range and launch date, and the Model 2, and the 4680 battery process, and...

Based on Musk's track record, you can pretty much count on this being a lot less than what is promised and a lot later.

I also just don't see the opportunity. I wouldn't call myself all that knowledgeable about WeChat and its ilk, but I think these "super apps" emerged as China's mobile revolution was taking off, meaning that people started out doing banking and ride sharing etc within these apps. In the US, all those services came out separately.

Admittedly the app landscape is fairly cluttered, but I just can't see the path to US consumers suddenly wanting to hail a ride inside of Xwitter. It's not how most Americans learned to hail rides, and the consumer value in having it all inside of Xwitter seems pretty minimal.

Side mirrors almost always leave a large blind spot directly behind and close to the vehicle. There's a reason that when firefighters are reversing their appliances they always have at least one of the crew physically get out and watch the area behind the vehicle.

Even a rear window and rear view mirror almost always leave a significant blind spot low and close behind the vehicle, which is why reversing cameras became a thing. When they're done well, they really are significantly safer, as well as sometimes making it a lot more reliable for most people to park the vehicle in difficult spaces.

One of the modern innovations I really would like to have is full AR on my windscreen. I want unexpected hazards highlighted in real time, particularly those that are more easily detectable by non-visual sensors, like big potholes or animals obscured by vegetation near the side of a country road. I want the actual driving line I need to take to follow my planned route through complex junctions overlaid slightly on my view of the road ahead. I want light amplification for night driving, ideally combined with some other technology that can reduce the glare from oncoming headlights to prevent dazzle.

Although I only want all of this if (a) it's implemented well and (b) any additional data it uses is reliably up-to-date and (c) there's an emergency shut-off that instantly clears everything off the windscreen in case anything goes wrong.

Don't worry. You probably have funky modern door handles that don't work when the power goes out anyway. Not that the power in an EV is likely to go out if it's underwater or on fire or anything.

Oh, don't be silly. Next you'll be making even more absurd claims, like that car theft was already a solved problem 20 years ago thanks to immobilisers, or that having separate physical controls for essential functions that you can find and use without taking your eyes off the road for several seconds to mess around with a touchscreen is safer, or that no-one ever hacked 100,000 cars at once from 1,000 miles away back when they didn't have always-on remote connectivity and allow OTA updates to their essential control systems.

Yes, as long as you're the one in the big, heavy car, it's great. Shame if you're the kid it's reversing over though.

Do you ever use reverse gear? What's behind you is pretty important when you're going backwards...

Yes, Apple's shared RAM model really works for them in the context of running LLMs locally. It's a huge advantage. As you say, not much use for those running other platforms, though.

I'm impressed that anyone can afford a new development PC in 2026. I'd need a second mortgage with the price of the RAM, SSD and GPU these days!

I'm not sure I fully agree. If you know what you're doing, LLM-based code can be quite helpful. I just built a Python scraper using Antigravity in a few hours that would've taken me many days of work and required a lot of effort to learn async function syntax.

It's not a super complex code base, so there's nothing architecturally complex about it. Even then, if I hadn't had a decent understanding of how playwright works, I would've had a much harder time debugging things and fixing some of the dumb decisions the LLM made.

The "anyone can build anything" hype is clearly bogus, but "people who can clearly specify things" can build a lot more than otherwise.