Copilot wrote me some code that totally does not work. I pointed out the bug and told it exactly how to fix the problem. It said it fixed it and gave me the exact same buggy trash code again. Yes, it can be pretty awful. LLMs fail in some totally absurd and unexpected ways. On the other hand, it knows the documentation of every function, but somehow still fails at some trivial tasks. It's just bizarre.
Fair enough, and that’s actually really good. You’re going to be one of the few who actually go through the trouble of making an account on a forum, ask a single question, and never visit the place after getting the answer. People like you are the reason why the internet has an answer to just about anything.
Interestingly, there’s an Intelligence Squared episode that explores that very point. As usual, there’s a debate, voting and both sides had some pretty good arguments. I’m convinced that Orwell and Huxley were correct about certain things. Not the whole picture, but specific parts of it.
Sure does, but somehow many of the answers still work well enough. In many contexts, the hallucinations are only speed bumps, not show stopping disasters.
That’s true. There could be a balance of sorts. Who knows. If LLMs become increasingly useful, people start using them more. As they loose training data, quality goes down, and people shift back to forums etc. Could work that way too.
"Some years ago, I provided my phone number to Google as part of an identity verification process, but didn’t consent to it being shared publicly."
That may have been the case at the time, but Google have a bad habit of updating legal documents and settings from time to time. Even if you didn't consent to it directly, you may have agreed to a contract you didn't read, which resulted in Google doing everything permitted in that contract. Chances are, the contract says that Google can legally screw around as much as they like, and you're powerless to do anything about it.
Based on the numbers from Purism, it could be a lot more than 25% more expensive to manufacture everything in USA. Purims Librem 5 costs 799 $, while the made-in-America version costs 1999 $. That's roughly a 2.5x difference. Obviously, economies of scale play a role too but let's assume that the same factor applies to iPhones too. If so, the fanciest iPhone would cost about 4000 $.
Zero, a notoriously ambitious hacker, had set his sights on a particular server. Rumors swirled in the dark corners of the internet that this server held a treasure trove: a database brimming with user accounts and password hashes. For Zero, breaching it would solidify his legend.
He spent days, then weeks, launching every exploit he knew. He tried SQL injections, brute-force attacks, phishing attempts, and probing for every conceivable vulnerability. Each time, the server remained unyielding, its digital defenses ironclad. Zero’s fingers flew across his keyboard, lines of code blurring on his multiple screens, but the coveted data remained tantalizingly out of reach.
Frustration mounted. Sleep became a luxury, and the thrill of the chase turned into a gnawing obsession. Yet, despite his relentless efforts and advanced tools, the server simply wouldn't yield its secrets.
Finally, after one last, exhaustive attempt failed, Zero leaned back in his chair, a bitter laugh escaping him. "Forget it," he muttered to his empty room, "that server is a honeypot anyway! Just a decoy set up to waste a hacker's time. There's probably no real data on it at all."
Oh, and the lesson. Almost forgot. Err... Don't be a noob, don't trust everything you read online, know what you're hacking... oh bugger this post is going off the rails. I'm sure there's a good lesson somewhere in there. Buried deep...
Take a really long rope and put one end on either pole of the Earth, and the other end on the equator. Use the shortest path, and make sure the rope is tight. No squiggles allowed! Chop that rope into exactly 10 000 000 equal parts. One of them is as long as a meter. Now you just need to find the right one.
Copilot wrote me some code that totally does not work. I pointed out the bug and told it exactly how to fix the problem. It said it fixed it and gave me the exact same buggy trash code again. Yes, it can be pretty awful. LLMs fail in some totally absurd and unexpected ways. On the other hand, it knows the documentation of every function, but somehow still fails at some trivial tasks. It's just bizarre.