pewter

Your analogy requires a powerful faction of people in Poland directly shooting rockets at Russian-occupied Ukraine. Still a significant event, but this descent continually shows the problem with analogies.

Fillory and Further by Christopher Plover

within

The Magicians by Lev Grossman

Even if some countries were to ban or heavily restrict AI tech, others will not.

Also, I’m convinced any country that heavily restricts its use will just be using their own at a federal level. All this untagged data that’s collected from us through surveillance is pretty valuable. If AI could help a government find a serial killer would they use it? If AI could help a government positively identify a revolution would they use it?

I think people would be surprised at how little hard drive space some models take up when you compare it to what they can produce. You can store LLMs on your phone. Cat’s out of the bag.

And despite any restrictions we try to make, there will be a company with great GPUs or TPUs that will be able to fine tune giant models without anyone knowing.

Yes, but it’s such a falsifiable claim that anyone is more than welcome to prove them wrong. There’s a lot of slightly different LLMs out there. If you or anyone else can definitively show there’s a machine that can identify AI writing vs human writing, it will either result in better AI writing or it would be an amazing breakthrough in understanding the limits of AI.

Post- high school I’ve seen explanations that describe magnetism as a consequence of electricity with the understanding of relativistic effects.

Google immediately jumped ahead in search when it started by having a simple webpage and using PageRank. This was a while before there were even Gmail accounts and all the tracking we’re given now.

At this point I’d settle for a search company that doesn’t care to track you, uses general (not specific) predictive search, implements Boolean search, and isn’t diminished in quality by SEO.

That last criterion is the hardest one. It might not even be feasible.

You’re basically forced to measure that as wealth. Otherwise we’re just pretending someone is poor just because they’re cash poor, yet I would argue that poor people have no real way to get 40 billion dollars to immediately lose money on a social media company.

The AI sphere isn’t like crypto’s sphere. Crypto was truly one thing. Lots of things can be solved with AI and modern LLMs have shown to be better than remedial at zero shot learning for things they weren’t explicitly designed to do.

If you have

1. lots of money
2. connections overseas
3. and an actual home in another country

I think it’s really hard to find an appropriate bail even if you aren’t witness tampering.

A lot of good suggestions in this thread, but for a combo of what you’ve already watched (Abbott Elementary and the British version of Ghosts) Bad Education is a comedy about a subpar British teacher.

Tesla and SpaceX seem very different.

Lots of companies are making electric vehicles. Tesla’s self driving software advantage is limited by people’s high standards of success for AI.

SpaceX, on the other hand, currently has no real world competitor in the reusable rocket space.

A lot of focus is being justifiably placed on the mom, but the kid’s actions very closely mirror allegations against his father, Malik Ellison.

The [father’s girlfriend] said she backed away to create space, but Ellison followed her, pushed her onto the bed, put his hand around her neck, and threatened to kill her, according to court records.

That attack wasn’t the only time the woman claimed Ellison was violent toward her, court records show. The woman called Newport News police again on Sept. 21, 2021 and told investigators that Ellison held a gun to her head and threatened to kill her, according to court records.

https://www.wavy.com/news/investigative/court-docs-father-of-richneck-shooter-assaulted-two-women-including-boys-mother/

This table only covers some of your criteria, but it’s pretty large. It’s a good source for finding another home instance.

https://github.com/maltfield/awesome-lemmy-instances/blob/main/README.md

Are e-cigarettes less harmful than regular cigarettes?

Yes—but that doesn’t mean e-cigarettes are safe. E-cigarette aerosol generally contains fewer toxic chemicals than the deadly mix of 7,000 chemicals in smoke from regular cigarettes.3 However, e-cigarette aerosol is not harmless. It can contain harmful and potentially harmful substances, including nicotine, heavy metals like lead, volatile organic compounds, and cancer-causing agents.

https://www.cdc.gov/tobacco/basic_information/e-cigarettes/about-e-cigarettes.html

AI is already better… than some people. A human using AI is probably better and faster at certain tasks than a somewhat skilled human is.

I bet midjourney is better at making concept art than the vast majority of the population.

I think we have a high threshold for success of AI. I saw a video a while back about how AlphaGo (an AI designed for playing Go) was able to beat a whole bunch of experts in Go. One expert used an atypical move and beat AlphaGo. People started reacting like “see? AI isn’t impressive. This genius beat it.” How many of us are geniuses? How often will geniuses beat better AI?

I’m wondering if there’s going to be a real problem when content gets dominated by AI and AI starts scraping their own hallucinations.