Typical tech-exec BS word salad.

If it were 20% more than the average Chinese made laptop, then maybe. If it's 20% more than the already vastly overpriced Macbook, then fuck no.

There, fixed that for ya.

I used Google's "AI" system a few months ago to have it make a simple reservation for three at a restaurant because I didn't want to wait on hold. It reported back that the reservation was confirmed and gave me all the info.

When we arrived they had absolutely no record of a reservation at that time under any name. Perfect.

I do not believe this even a little. In my experience call center workers do not like AI and customers hate it so much that they are now asking support workers to prove they are human, like some real-time voice captcha puzzle, because they refuse to deal with it.

See also: https://www.techspot.com/news/...

Deciding that "general intelligence" can be determined by how many billions in profits something makes is why some idiots think Elon Musk is smart.

They were saving up their limited funds for a Switch 2. Note that older gamers with a more reliable income aren't represented. Also - a four week period doesn't represent a trend. Come back when you have 12 months of data and we'll talk.

We'd have to see whatever license agreement they agreed to when they last installed an update to the software, which could have come with new terms allowing this. If they refuse the submit to the audit, VMWare might be able to remotely kill all the software, since I'm sure it's had some kind of online component to the licensing for years now.

My wife's new Motorola phone told her that it was using AI to adjust how it charges her battery. I'm 95% sure this isn't true - it's just using the same algorithm that Android implemented years ago to track when you tend to plug your phone in at night and when you wake up so it can charge as slowly as possible to maintain battery health.

This seems to be part of a larger trend of companies labelling every type of automation as "AI" because that's the buzzword du jour.

Most of the Big Companies across America are increasing their workforce

Four in five S&P 500 companies now employ as many or more people than they did a decade ago.

I, for one, welcome our new quadrupedal badminton-playing overlords!

All good questions that I'm sure will come up in the lawsuit.

The reason I said the question of specific infringement is "more compelling" is because I think the "theft" question is sort of abstract and might not get very far ultimately. Whereas any IP holder could potentially just go after genAI with a massive "infringement sting". Pay a subscription fee and start prompting in every way you can to get it to generate loads of infringing content, and then sue them for actual infringement.

If the AI companies want to claim that their training is no different than a human reading books and studying art, then treat them like a human that you paid money to create a work of art. If you hired an artist and paid them to make you a bunch of paintings of Disney characters, and they did, would that artist be liable for infringement? If they did it a lot, over a period of time, would that be grounds for a suit larger than simple damages?

I don't know, IANAL.

True, but it gives control of what others do with the material after they read it. For example, if I memorize a copyrighted short story or news article, then I type it into my computer with a couple words changed, post it online, and sell ads against it, that's not "fair use" is it?

This is basically what newspapers and authors have proven genAI can do - using prompts to basically recreate book passages and article almost verbatim.

It's the same thing - monetization of other peoples' works without permission.

Exactly my point. The business model isn't sustainable in its current form in a legal way - hence them being sued over and over. It's about time.

If you use an SaaS tool that you pay for and then simply ask it to generate a spline that "looks like a nike swoosh" then it might be infringing, because the company would have to have programmed the tool to a) know what "nike swoosh" means, and b) to recreate something resembling it. If the tool had never seen the logo and hadn't been trained to associate it with a brand, then it wouldn't be able to generate what you asked for.

There's a huge difference between a blank canvas where you as an artist have to move a pointer around in a path that replicates an IP, and a tool where you simply type a few descriptive words and it creates the image for you.

If you want to take the analogy to a ludicrous degree, there are some "tools" that are deemed illegal or inappropriate for the public market because despite having "legitimate" uses, they make criminal activity too easy. Lockpicks and signal jammers are not allowed in many places, automatic weapons are banned almost everywhere, and color printers that can perfectly duplicate money, for example.

I'm not saying Midjourney is a "dangerous weapon" but in their current state it's a tool that can very easily facilitate infringing activities on a mass scale, so there's precedent to look at these issues from a legal standpoint.