HI capital (with his connections in particular) will probably get the returns where it already is - building out AI infrastructure (starting with making chips) for which is this is unjustifiable demand still for the next little while anyways.

The spy is more worrying. US Gov and a lot of other big customers would not find that funny at all. I'm still surprised though... he made so much $$$ for everybody. This could be the black swan that brings the long-awaited AI market collapse.

Yeah, if every other airline makes theirs ai friendly. United will not get those bookings.

Counter point, does united want those bookings?

I don't' know. It looks like they are afraid that these are bargain hunters who would at best buy a ticket near cost for them. And they make most of their money off of branded credit cards.

Counter point, right now people using AI to book travel are sophisticated and on the wealthier side of things. I'm guessing United is just stepping back to see how they can better monetize ai agents use of their systems. Right now they can't show them adds or anything the end user will see, but maybe they could create a MCP server for dealing with them that would also offer deals including that credit card sign up promotion that actually earns them money.

All of summer vacation is about farming, originally

Unless the award is carrots. £1 / lb, all the time.

Energy is at some point causal, but that doesn't have to mean oil in all cases.

What it is is a possible contender to claw away at NVIDIA's history-making margins and profits, slashing AI compute cost and resources through efficiency gains. That's a big deal.

Pfft, buncha Manga-adjacent animation fans turned around 360 degrees by the anti-AI fad. AI, it's a form of animation. If somebody decided Zombee Lovers 2 was going to just be straight CGI or hand animation nobody would say the cast was "replaced" by animators.

To that point, I used to spend a fair amount of time on salon message boards. It was around the height of gay rights around maybe 2005? I remember scrolling down their page, every single 'story' was on gay rights... I checked, one after another... gay... gay, gay, gay. 30 stories in a row. I'm sympathetic to the issue and not averse to the stories but salon.com actually ceased to function as a general news and politics site.

All news outlets report mostly true facts. The most subjective and important act in writing news or history is deciding which facts are 'newsworthy' and deserve reporting on. Salon decide once and for all what side to take on all future issues a long time ago.

"This is why people who do real work don't take kindly to urban midwits telling them they could be forced to buy electric vehicles instead." I feel increasingly that way about taxpayer-funded oil wars. It doesn't seem like oil needs to be this huge, world-controlling thing of the world any more. With any EV you plug in you car, it drives around, no more gas station any more. It's great. I still use my SUV for trips and towing but those are luxuries, I do because I want to, I don't have to.

I have been using ScreenFloat (and similar tooling) for decades; those are more useful than this because it works on every application and not just the browser.

It's quite possible I am missing something from their demo, but based on what they show: 1) it's just for Firefox; and, 2) it's really kind of a pain in the ass to use.

Repost with paragraph tags:

I wouldn't mind a required label for it, like the explicit lyrics warning on CD's. Its laguhable to think AI isn't used in games. What are they going to do, have an actor in India mechanical turk every pretend person? That was never a thing. "Make 1000 varying zombie skins" is a fine job for AI.

Who are you going to pay to sit around and draw up 1000 NPC's? Of course there's AI in games.

https://ieeexplore.ieee.org/do... [ieee.org] Abstract from 1990: An experiment was conducted where neural networks compete for survival in an evolving population based on their ability to play checkers. More specifically, multilayer feedforward neural networks were used to evaluate alternative board positions and games were played using a minimax search strategy. At each generation, the extant neural networks were paired in competitions and selection was used to eliminate those that performed poorly relative to other networks. Offspring neural networks were created from the survivors using random variation of all weights and bias terms. After a series of 250 generations, the best-evolved neural network was played against human opponents in a series of 90 games on an Internet website. The neural network was able to defeat two expert-level players and played to a draw against a master. The final rating of the neural network placed it in the "Class A" category using a standard rating system. Of particular importance in the design of the experiment was the fact that no features beyond the piece differential were given to the neural networks as a priori knowledge. The process of evolution was able to extract all of the additional information required to play at this level of competency. It accomplished this based almost solely on the feedback offered in the final aggregated outcome of each game played (i.e., win, lose, or draw). This procedure stands in marked contrast to the typical artifice of explicitly injecting expert knowledge into a game-playing program. Published in: IEEE Transactions on Neural Networks ( Volume: 10, Issue: 6, November 1999) https://ieeexplore.ieee.org/do... [ieee.org]'

I wouldn't mind a required label for it, like the explicit lyrics warning on CD's. Although its laguhable to think AI isn't used in games. What are they going to do, have an actor in India mechanical turk every pretend person? That was never a thing. Who are you going to pay to sit around and draw up 1000 NPC's? Of course there's AI in games. https://ieeexplore.ieee.org/do... Abstract from 1990: An experiment was conducted where neural networks compete for survival in an evolving population based on their ability to play checkers. More specifically, multilayer feedforward neural networks were used to evaluate alternative board positions and games were played using a minimax search strategy. At each generation, the extant neural networks were paired in competitions and selection was used to eliminate those that performed poorly relative to other networks. Offspring neural networks were created from the survivors using random variation of all weights and bias terms. After a series of 250 generations, the best-evolved neural network was played against human opponents in a series of 90 games on an Internet website. The neural network was able to defeat two expert-level players and played to a draw against a master. The final rating of the neural network placed it in the "Class A" category using a standard rating system. Of particular importance in the design of the experiment was the fact that no features beyond the piece differential were given to the neural networks as a priori knowledge. The process of evolution was able to extract all of the additional information required to play at this level of competency. It accomplished this based almost solely on the feedback offered in the final aggregated outcome of each game played (i.e., win, lose, or draw). This procedure stands in marked contrast to the typical artifice of explicitly injecting expert knowledge into a game-playing program. Published in: IEEE Transactions on Neural Networks ( Volume: 10, Issue: 6, November 1999) https://ieeexplore.ieee.org/do...'

It's this moral panic over water and power for datacenters - results in people going too far to prove a point.