I’m not an expert. I’d just expect a neural network to follow the core principle of self-improvement. GPT is fundamentally unable to do this. The way it “learns” is closer to the same tech behind predictive text in your phone.
It’s the reason why it can’t understand why telling you to put glue on pizza is a bad idea.
Calling GPT a neural network is pretty generous. It’s more like a markov chain
Babe wake up, new market manipulation just dropped
From the robot’s perspective, this article is horrifying.
Assembly is ancient Egyptian heiroglyphs.
deleted by creator
I’ve been a professional programmer for nearly a decade and I just realized that C# is C++++ with the pluses stacked
This might just be my computer-focused life talking
I’m a software eng too, but I have broad interests. Like I said, the philosophic use doesn’t really have a place in this discussion and I messed up by bringing it in. The only way it would be relevant is if the universe is a simulation because, as you guessed, then free will itself becomes part of the equation.
I also don’t know why predictability would be solely based on the numbers that came before
There’s a miscommunication happening here, and I’m wondering if I’m not explaining myself well. Election predictions use polling as their dataset, and there are no calculations that really go into predicting the results other than comparing the numbers within those sets. That’s why they’re notoriously garbage (every single pollster had Hillary winning in late October 2016, for example). Also, there aren’t any calculations that go into a CEO/Boardroom’s intuitions on how shareholders will react to policy changes, so I’m not sure about the relevance here. In the case of pi, there is no dataset that you can use that tells you what the next unknown number in pi is. The only way to get that number is to run a very complex calculation. Calculations are not predictions.
As I said, you can’t predict the next number simply based upon the set of numbers that came before. You have to calculate it, and that calculation can be so complex that it takes insane amounts of energy to do it.
Also, I think I was thinking of the philisophical definition of “deterministic” when I was using it earlier. That doesn’t really apply to pi… unless we really do live in a simulation.
There’s no way to predict what the next unsolved pi digit will be just by looking at what came before it. It’s neither predictable nor deterministic. The very existence of calculations to get the next digit supports that.
Note: I’m not saying Pi is random. Again, the calculations support the general non-randomness of it. It is possible to be unpredictable, undeterministic, and completely logical.
Note Note: I don’t know everything. For all I know, we’re in a simulation and we’ll eventually hit the floating point limit of pi and underflow the universe. I just wanted to point out that your example doesn’t quite fit with pi.
Yeah, but your number doesn’t fit pi. It may not have a pattern, but it’s predictable and deterministic.
deleted by creator
If it changes the “entire world”, I would very much prefer it not to change the world for the worse, but that’s the current trend.
This could run in Javascript if you setup print as an alias for window.alert or console.log
Yeah! What color do you think it should be? I’m thinking red.
Yes, IBM is the parent company, but I was specifically talking about Red Hat buildings
Red Hat has multiple large buildings around the world, but that doesn’t fit the meme.
You’re right, my bad.