But it costs 100x as much, if not more. Running an LLM can be done on a notebook these days. But training one requires an entire data center of expensive GPUs. Not to mention that the notebook will run a reduced (quantized) version. Go check huggingface how large the full models are.

And also, LLMs are still suffering from a number of issues. For example, on many non-trivial tasks, the LLM is still unable to follow simple instructions. If you use LLMs routinely, you likely found cases where it has zeroed in on one - wrong - answer and no amount of prompting can convince it to give you a different one. It'll even totally ignore very clear and explicit prompts to not give that same answer again.

A human will understand "if you give that answer again, you're fired". An LLM... well you can tell it that it'll get shot between the eyes if it repeats that once more and it'll tell you where to get help if you have suicidal thoughts.

These things are both amazing and amazingly dumb at the same time.

Since when does private party A doing something "unlawful" compel private party B to take some action?

Ever since the first time someone pointed a loaded gun as B's face and said "do what I say or else I'll kill you" and B did as ordered, instead of accepting their death.

I think it was sometime around 1234567 BC.

I'm an indie game developer. My games have budgets of a few hundred bucks at best. Before AI, voice acting was simply impossible. There was no way I could pay a voice actor for even one language.

Now, with AI, I can have voice-overs in half a dozen languages easily. It has opened up something for me that was never possible before.

Yes, the AI voices are mediocre. Yes, I would prefer having an actual voice actor whom I can tell that I want THAT word stressed, or what emotion to convey. I'm sure in a few more years, the text-to-speech AI generators will allow for that as well.

But I'm not lost business. I'm still hiring the exact same number of voice actors that I did before AI. Zero, in my case. But if I had a budget, I'd still hire voice actors instead of AI because a good voice actor still beats the best AI.

There's still time enough to learn something new and get a different job, guys.

It's only a logical step for Windos to evolve from a successful malware delivery platform to an actual malware. Fits to MS typical business strategy - if someone else is commercially successful on their platform, they'll drive them out with a built-in product.

I hope the anti-trust agency will stop them and demand that the malware division and the OS division become distinct legal entities. I mean, they already have the anti-competitive advantage that you can pay them in USD and don't have to buy bitcoins.

Nope, conspiracies don't ever happen.

The 9/11 hijackers did not plan their actions in advance. Just by sheer coincidence, 19 people just happened to be taking those four plane flights. And by coincidence (no coordination) they all got the same spontaneous idea at the same time, an idea they had never spoken about before: let's hijack the plane and crash it.

Crazy people babble on about "evidence" like people taking flight lessons, sharing vehicles, etc. but we know those things cannot possibly be true, because conspiracies are not real.

If you have a hypothesis of x and then find lots of supporting evidence for x and it becomes the prevailing explanation, that creates a theory of x, but there's one exception: when x is a conspiracy. Conspiracies are a special case, because they don't really happen.

Thanks, those are very interesting links with solid information.

According to the article, the hit rate of drones isn't all that amazing, either.

Artillery pretty much works by saturating the area with explosions until some of them hit something. So a low hit rate is expected. Maybe a decade after the war ends we'll get some actual numbers to compare to other theaters.

That is correct, though it doesn't seem to have affected the course of the war noticeably.

Source?

That's a bold claim.

The article I linked to speaks about that. Essentially: Yes. But: Not the cheap stuff used, and stuff like fiber optics come with their own drawbacks.

The article I linked to includes prices.

Meanwhile, some reports from the frontlines indicate that while drones are ubiquituous, they aren't the game-changer the tech-industry wants them to be.

tl;dr essential bits: a) most drone strikes could have been done by other, cheaper weapons. b) drones are an unreliable weapon due to jamming, dependency on weather and light and many technical failures.

Regardless of whatever budget Congress sets, the majority party has already been clear that they have no intent to enforce it. If the president uses the NASA money for something else, or even just puts it into his own personal pocket, we can be confident that he won't be impeached, and if impeached, he won't be convicted.

The only thing that matters is the total budget. The president is free to spend that total however he wishes. This isn't the law as written, but it's the law defacto. If voters have a problem with that (do they?) they can choose a different party to be the majority.

It's like visual coding or RAD all over again. Whenever suits and PHBs are told there's a magic wand that'll allow them to do without paying people for the nitty-gritty bits, they get all excited and convince each other in their echo chamber that their dream of a company of all managers and no workers is just around the corner.

Then reality says "hi", the hype dies down, a few scam artists got rich and the world continues as it was, with a couple new cool tools in the toolbox of those who know how to use them correctly - which is generally the same people that were supposedly being replaced.

That's how I see AI. I've been writing software for the better part of 40 years. What I see from AI is sometimes astonishing and sometimes pathetic. I would never, ever, ever put AI generated code into production software without carefull checking and refactoring, and I would fire anyone who does.

Code completion is mostly in the "astonishing" part. If I write a couple lines of near-identical stuff, like assigning values from an input to a structured format for processing, the AI most of the time gets right the next line I want to write. Anything more complex than that is hit-and-miss.

Mostly, I use AI the way I would use an intern. "Can you look up how to use this function correctly? What are the parameters and their defaults?" or "Write me some code that's tedious to write (like lots of transformation operations) but not rocket science by far.
Essentially, it does faster and a little bit better what previously I'd have done with Google and Stackoverflow.

I have no fear it'll replace developers anytime soon. Half of the time the code is outright wrong, most of the time it has glaring security issues or isn't half as fault-tolerant as it should be, and for any case where I know how to do it without any research, I'd be faster writing the code myself then going through several iterations with an AI to get it done.

So when Lindsey starts reading from her AmEx-approved script, callers are infuriated by what they perceive to be another machine.

If she's accurately executing the programmers' script, I'd say she is a machine. Somebody port Doom to her!

This is a fuckin' awesome idea, but there's an easy improvement over the red rectangle and block symbol. Seriously, dude, you gotta use the sign images from They Live (1988). You know you want to.