Respectfully, I think we cracked basic intelligence. What do you imagine under basic intelligence?
LLMs can do homeworks, pass standardized exams, give advice WITHOUT ANY SPECIFIC TRAINING.
You can invent an imaginary game, explain the rules to the LLM and let it play it. Just like that.
You can invent an imaginary computer language, explain the syntax to the LLM and it will write you valid programs in that language. Just like that.
If that is not intelligent I do not know what is. In both cases, the request you put in is imaginary, exists only in your head, there are no previous examples or resources to train on.
> Respectfully, I think we cracked basic intelligence. What do you imagine under basic intelligence?
It all depends on your definition of intelligence. Mine is the ability to solve novel problems.
AI is unable to solve novel problems, only things it has been trained against. AI is not intelligent, unless you change the very definition of the word.
I challenge you to imagine an imaginary game or computer language, explain the rules to the LLM. It will learn and play the game (or write programs in your invented language), although you imagined it. There was no resource to train on. Nobody knows of that game or language. LLM learns on the spot with your instructions and plays the game.
I cannot understand grad school level mathematics even if you give me all the books and papers in the world. I was not formally trained in mathematics, does that make me not intelligent?
"Ability to solve novel problems" does not mean ability to solve all problems, nor to have all the knowledge in the world. A caveman can be as intelligent as you and I, even without being able to read.
Not having knowledge in mathematics has no impact on your intelligence quotient.
I agree. But LLMs do solve novel problems in that case, you just have to explain it to them like you would have done to an intelligent caveman or me.
Which novel problem can't an LLM solve? I gave the example of an imaginary game because the LLM cannot have possibly trained on it as it is imagined by one person and nobody knows how to play it.
Huh? It's a meme that LLMs can't follow the rules of chess. Just tried tick tack toe on GPT 3.5 and not only did it pick bad moves it also failed to evaluate the win condition.
If LLM could invent consistent imaginary games (or anything, like a short novel, or a 3 page essay on anything it want), maybe i would agree with you. The issue is that anything it create is inconsistent. The issue might be an artificial limitation to avoid copyright issues, but still.
But even that. Did you try to use GPT4 as a chess engine? I have issues with the Slav defense when i start with the queen's gambit, i tend to loose tempo or position, or both. I asked him continuations, and it was either wikipedia entries or nonsense, no in-between, no interesting insight. Now, i have asked a regional champion a bit before that (he is around 2.2k elo, so not exceptionally good) and although i can't seems to understand or use the concepts, he gave me interesting enough ideas to build on it.
Not saying that chatGPT isn't a great tool to write documentation or fiction (half my TTRPG campains are featuring description by ChatGPT), but i wouldn't call it intelligent.
Chess is a very specific field that requires training. Chatgpt may not be optimized for chess.
And I think chatgpt has some issues visualizing stuff like a chess board.
Therefore to get a decent answer you'll have to explain that you are a professional chess player. You'll have to describe what tempo and position means in chess. You'll have to describe what a gambit is etc. After these steps it will understand and guide you in whatever you need.
If you succeed you can release it as a customGPT.
It's a bit like asking a tea from the ship's supercomputer in hitchikers guide to the galaxy.
Consistency, for one. I have asked LLMs the exact same question twice in a row and got wildly different answers. Intelligence presupposes understanding. When I ask an LLM “give me the first X of Y” and it replies “I cannot give you the first X of Y because there have only been X+10, here’s the first X+5 instead”, I’m hard pressed to call it intelligent.
Have you tried specifying you field of inquiry which was algebra. Try saying solve this equation for me. I am a lawyer by day so I constantly face limitations of natural languages. The solution is to write less ambiguous prompts.
The field of inquiry was not algebra. It was a straightforward question using real numbers. I asked it about the first <number> kings of <country>. I don’t recall the exact number, only the stupidity of the answer.
So you understand, let’s say I asked for the first 20 kings of England. It told me it could not give the first 20 because there had only been 30, and that it would give me the first 25 instead.
I gave minimum context like this: "I have a history exam. You are an expert in British royal history. List me the names of 20 kings and queens in England."
The answer was: "Certainly! Here's a list of 20 kings and queens of England:
1. William the Conqueror
2. William II (Rufus)
3. Henry I
4. Stephen
5. Henry II
6. Richard I (the Lionheart)
7. John
8. Henry III
9. Edward I (Longshanks)
10. Edward II
11. Edward III
12. Richard II
13. Henry IV
14. Henry V
15. Henry VI
16. Edward IV
17. Edward V
18. Richard III
19. Henry VII
20. Henry VIII"
I disagree. They are not just text generators. LLMs are increasingly being multimodal they can hear and see.
We humans are also text generators based on text content. What we read and listen to influences what we write.
Llms are intelligent at least as us humans, they can listen, read, see, hear and communicate. With the latest additions they can also recall conversations.
They are not perfect. Main limitations are computing power available for each request and model size.
Have you tried Claude Opus 3 or GPT 3.5 or Gemini?
Microsofts copilot is dumb (I think they are resource constrained). I encourage everyone to try at least the 2-3 major LLMs before giving a judgement.