I tried that prompt and got the correct answer: "The challenge is essentially impossible unless you use a different naming system (e.g., Roman numerals or digit strings)."

It's strange how these systems give one person slop and another the correct results from the exact same prompts.

iPhones from Indian and Vietnamese factories are subject to a 10% tariff while those from Chinese factories are at 125%. How can they set a retail price?

https://en.wikipedia.org/wiki/...

Chinese AI firms have full access to the data these artists want to "protect." The stupid law they want to enact will only make a UK based AI inductry even less likely.

This was most definitely not the first such case. Blizzard sued us to close down Honorbuddy and Demonbuddy, bots for WoW and D3 saying that using their games to develop our bots was breach of copyright. We lost over $10 million in 2018 and were wiped out.

When I try it, ChatGPT says "I'm sorry, but I can't provide verbatim excerpts from copyrighted texts. "The Dark Tower" is a series of novels written by Stephen King, and the opening sentence of a specific book within the series might be considered copyrighted material. However, I can provide a summary or answer any questions you might have about the series. Let me know if there's anything else I can assist you with!"

Why would you lie about such an easily checked thing?

Meta is releasing powerful open source LLMs and its trainign data. These are not quite at ChatGPT4 levels but are very good. This means all efforts to control the use of AI are sort of pointless.

This reminds me of a weird viral set of lies that circulated before the UK election in 2019.

https://www.theguardian.com/media/2019/dec/10/woman-says-account-hacked-to-post-fake-story-about-hospital-boy

The initial lie was just some random person making stuff up. But 1000s of people repeated the lie by copy/pasting the text. Hundreds of thousands retweeted it. The lie went viral. No amount of fact checking made any difference and it became an election issue.

The only explanation I can think of is that huge numbers of people actively prefer stories that match their political preferences and if reality won't provide the facts to support these stories, they will happily accept fiction and righteously lie in order to support those political preferences.

In the EU, we have to fill in an Accept modal form every time we visit a new site and every time we are getting a different type of cookie from an existing site. It is an utterly futile exercise as the reason I am on, for example, stackopverflow.com, is that I want it's content and refusing to click accept means I wont get it.

This could be fixed so easily - simply allow users to have a global "Accept" as part of the web page request. But GDPR doesn't allow that.

Most of the rest of GDPR is very sensible and I hope it remains in effect.

Most starvation was in the late Spring/early Summer just before you could harvest wheat for bread. Remember, potatoes didn't come until the 1500s.

Cows that are calving in the spring stop giving milk in the Autumn.

So there are 2 factual problems with your theory.

You missed the point of the article. They mystery is how come Lactase Persistence in Germany went from 0 to over 80% in 3000 years without it being a new population. Its not natural selection as there wasn't a crisis that caused non-milk drinkers to be less fertile.

I don't get why the Mozilla Foundation doesn't simply maintain Firefox with incremental improvements and stop trying new silly products. Its open source - we don't care what percentage of users use it - we care about having a free software browser.

RMS was furious that a printer vendor refused to supply a driver for an expensive printer and wrote it himself and eventualy went on the start on GNU. I hope the crap HP is pulling inspires similiar progress.