It's a terrible definition of intelligence. Intelligence is far more than just predicting things based on a massive set of similar things. We do do pattern recognition, but I don't need to have seen 10k dogs to be able to recognize a dog. And it's not a difference of degree, either. "Meaning" is something I am able to derive from my experiences, and it is not something an LLM does or will ever be able to do (nor is training data even analogous to having experiences in any way shape or form.)
I didn't say anything about training methods. I said an intelligence can predict things accurately, most particularly if it can predict the outcomes of possible actions it can take then it can do planning. Its especially important that it should be able to predict outcomes in cases that are not exactly the same as things it has seen before, but how it gets to that capability level is not important for understanding what the capability unlocks.
This is one of the most distinctive qualities of human intelligence compared to many animals - is our ability to adapt to new situations and make accurate predictions of the outcomes of our actions. Other animals can also do this but over very narrow time horizons and situations.
AI agents are better at it than any animal, and better at it than humans in some domains.
Is "meaning" inherent to intelligence? Is emotion and the ability to have a subjective experience inherent to meaning? Genuine questions that I'm not sure we have good answers for yet.
Your other point that humans need very few examples for learning vs. what LLMs need is an interesting one. A few researchers talked about this very thing on the most recent episode of Dwarkesh's podcast.
My own view is that it seems unfair to compare an LLM's training with just what a human gets over the course of a lifetime, because our brains have been trained by a billion years of evolution for pattern matching. I'm not at all confident that AI models won't catch up.
Oh, I don't know, let's math it. I keep hearing that women need to be twice as good to make it half as far. Sooo, 20% women in STEM x 2 x 2 (twice as good; and, multiply by two to make up for half as far) = ... 80%
I keep reading this positive stereotype about women needing to be twice as good to make it half as far, and yet this outcome would be a the natural consequence of that if assessment was to suddenly become fair. People need to better choose their 'ladies in the workplace' stereotypes methinks. Unless it's true, in which case...
Let's see, 20% Women in STEM x 2 x 2 (twice as good; and, multiply by two to make up for half as far) = ... 80%
Answering this question feels like the hardest problem I've solved yet... ;) Because, I don't know: I've never really thought "this one! THIS is the hardest!" You just iterate and things get more and more challenging as you build skills. What seemed hard to a junior tech doesn't seem hard to me as a senior tech now. It's all just engineering. It is all just sitting down, reading manuals or prior art, getting familiar with protocols or fundamentals, and building maps in your head until you understand something. Then building proofs of concepts and outlines; then, applying a bunch of troubleshooting principles; repeat until problem is solved. I've written academic papers this way, I've built streaming servers off esoteric industrial process-control database APIs, I've done process visualizations, I put a model railway online (before that was an out of the box thing)... and it's all the same: use what other people did, understand it, and then build from there.
> "That wasn't so hard. Why did it take me so long? Am I bad at this stuff?"
Hah. Always. Hindsight bias and impostor syndrome are a fun mix! I remember writing a blog suite (with comments!) in Perl in the late 90s; back then, without S.O. and other knowledge-sharing beyond some Usenet forums, inventing the wheels as we went along... it was all hard.
I coded the same thing (actually an online magazine with comments on articles, a forum, and a form for signing up for email updates), around the same time. I used classic ASP. I shudder to think of all the security holes I must have had.
You do realize that finance is one of the worst places for representation of women and is recognized as a cesspit? (http://www.forbes.com/forbes/2009/0316/072_terminated_women....). It's wonderful that your friend has had positive experiences throughout her career.
She never said finance was sexist, only that people were rude in general. I'm not making any claims myself, only relating the experience of a colleague.
Found this sneaky advertorial for CodeCore Bootcamp, which promises to make you a software developer in only 8 weeks! 8 weeks! Sos that you can access fabulous jobs in Vancouver for the exceptionally high average salary of $50k a year! FIFTY K, PEOPLE.
So. 8 weeks? To make a software dev? A Hootsuite intern, perhaps...
I'd argue that it takes several years to develop a problem-solving toolkit, best-practices, and just enough fundamental theory to be a half-decent developer with both breadth and depth enough to write non-embarrassing, maintainable code without needing tonnes of support, feedback (which is always helpful, but particularly essential in early days), and nights spent falling asleep with K&R on your face. Whether that's in school or out, getting employable takes time: time spent doing code, and thinking about it, and talking about it. I feel like these CodeGuru guys are being misleading at best, and that article reads to me almost like satire.
reply