Damn, even with inflation? You must be super on guard. That being said. I will turn on my A game to shave $0.50 off of a set of coasters from a tourist shop while on vacation.

In terms of practical commercial uses, these highly human in the loop systems are about where it is and there are practical applications and products build off of it. I think what was sold though is a much more of either a replacement of people or a significant jump in functionality.

For example, there are products that will give you an AI summary of a structured or fairly uniform document like a generic press release, but there’s not really a good replacement for something to read backgrounds on 50 different companies and figure out which one you should invest in without a human basically doing all of that work themselves anyway just to check the work of the AI. The latter is what is being sold to make the enormous cost of hosting and training AI worth it.

This is awesome, and done by some really talented kids who are clearly smart motivated and willing to put in the work to get this project out the door.

Now imagine if the resources to do this kind of work as well as the background education and things like food security and economic stability were given to kids outside of an exclusive private boarding school? We’d empower some of the most imaginative of us to accomplish so many more amazing things.

The issue is AI is just too broad of a term. It’s also not a magic bullet and comes with its own problems so it’s not even the best tool for the job many times.

No, Google and Amazon were actually well run businesses with sensible business plans to meet needs in the market and did it well.

It’s easier said than done. A few key pieces took decades to figure out and even now many can only be produced by one or two companies, like ASML.

I mean this is criminal negligence that lead to many deaths. I’m not a fan of the death penalty but they should be held as liable as a drink driver or someone who skimps on building materials and kills someone when the building collapses.

Regardless of what one might think should happen or expect to happen, the actual psychological effect is harmful to the victim. It’s like if you walked up to someone and said “I’m imagining you naked” that’s still harassment and off-putting to the person, but the image apps have been shown to have much much more severe effects.

It’s like the demonstration where they get someone to feel like a rubber hand is theirs, then hit it with a hammer. It’s still a negative sensation even if it’s not a strictly logical one.

Careful, sounds like that one might foam up out of the bottle.

He brought shareholders value by getting people pumped to buy the car.

From the perspective of his daughter who knew George Carlin personally, I can see how this would be disturbing. It’s as if someone strung up a dead relative like a puppet and put on a show.

I think in more abstract terms from someone who just saw his standup, it’s a fun novelty as long as they’re not profiting from it or misrepresenting it.

Oh no, my local community events aren’t announced on Lemmy, therefore I should just never interact with people in my town and spend more time browsing lemmy.

If you only allow all or nothing adopters, you’re going to end up with nothing.

Many sites I visit are not because I like the design of the site, but that I like the content that the users it has attracted have uploaded. Sure, there’s YouTube alternatives, but they’re not a replacement for the scale of YouTube. There’s still a few subreddits I read because they don’t have a good lemmy alternative.

I use uBlock on Firefox and never got any kind of notification either. Not a paying customer.

Haven’t read the paper so not sure about the specifics, but if it relies on subtle changes, would rounding color values or down sampling the image blur that noise away?

Bold of you to assume we’ll still have fish after global warming, oil spills and micro plastics do their thing.

There’s a difference between being able to technically do this, and deploying it as a product with legal checks and scaled up infrastructure.

Since the article was about generative voice models, I assumed the ad part was also about that, but you’re right it might just be using voice recognition.

It was until you hit the part in the article about using AI voice generation to target ads. When election season rolls around, we’ll get all sorts of clips of people endorsing Trump without their knowledge.

AI is a tool, its value is dependent on whatever the application is. Transformer architectures can be used for generating text or music, but they were also originally developed for text translation which people have fewer qualms with.

I’m kind of surprised people are more concerned with the output quality for chatGPT, and not where they source their training set from, like for image models.

Language models are still in a stage where they aren’t really a product by themselves, they really need to be cajoled into becoming a good product, like looking up context via a traditional search and feeding it to the model, or guiding it towards solving problems. That’s more of a traditional software problem that leverages large language models.

Even the amount of engineering to go from text prediction model trained on a bunch of articles to something that infers you should put an answer after a question is a lot of work.