There is no such thing as a Stupid Question!
Don't be embarrassed of your curiosity; everyone has questions that they may feel uncomfortable asking certain people, so this place gives you a nice area not to be judged about asking it. Everyone here is willing to help.
Reminder that the rules for lemmy.ca still apply!
Thanks for reading all of this, even if you didn't read all of this, and your eye started somewhere else, have a watermelon slice ๐.
That's one of the "benefits" of Character.AI.
The only thing I found it good for was playing D&D by making the AI characters my PCs. The fact it would fuck shit up all the time just made it feel like playing with real people that also don't fully understand the rules or interpret the rules different.
A conversation/collaboration... not really.
You can create a 'swarm' of agents with differing roles, define different roles and phases, to have it iterate on a problem.
Groups of agents of the same role, operating in parallel, should ideally be using different models (or have context that gives them differing goals - eg focused on maintainable abstractions, security, scalability, test case identification, etc).
The implementation can do a similar thing - a code generator followed by reviewers, proposals for action, and then apply improvements... and you can iterate on testing or benchmarking too, all before hand-over.
This can improve results (at a non-trivial cost sometimes, so budgets are important) and it will still miss sometimes. You can help it of course with hints, directions or even implementations or stubs of implementations of abstractions you expect.
I love your reply and I just want to add for @cheese_greater@lemmy.world that the llms don't understand their output or your input or in this case the input and output of the various "conversing" llms. eseentialy they can't converse or colloborate in the way we do but what @taldennz@lemmy.nz has is right and being used by at least some of the models although I assume by now all are doing something like it.
From what I've read, the multiple AIs generally do have a result, just not a good one.
https://www.sciencedaily.com/releases/2024/03/240318142438.htm
But perhaps someone has a link to positive outcomes.
what you're proposing requires them to reason and understand each other. LLMs don't do that, they take text input and construct an output based on words (tokens) that they have mapped to be close to the ones you entered into your prompt.
it's a clever way to produce a plausible response, but it's not thinking or reasoning.
I'm not sure most people are thinking or reasoning either ๐
Didn't you argue that deep learning models were able to think like humans and then used this exact line in your arguments? Like a month or two ago?
Absolutely. It is not thinking in the same way we do.
Putting aside the planning orchestrator and focusing just on the LLM.
The agent can do this in stages to try and decide what the complete set of input tokens should be, and at what point to stop trying to get more output tokens.
You can use the orchestrator approach to then try to get other models to validate the outcome and refine it - but it's all just prodding the statistical model.
This talk was interesting. I'm a lot less enthusiastic on the topic than the speaker... but this is closer to how I think the industry can see a net gain from AI - before the slop-errors from taking expertise out of the loop hits critical mass.
