There’s already more than enough training data out there. The important thing that remains is to filter it so it doesn’t also include humanity’s stupidest data.
That and make the algorithms smarter so they are resistant to hallucination and misinformation - that’s not a data problem, it’s an architecture problem.
Stupid data can be useful for training as a negative example. Image generators use negative prompts to good effect.
Butbutbut my ignorant racism is the truth!! That’s why I hear it from everyone, including [insert near by relatives here]!!
Well is the goal truth? Or a simulacrum of a human?
Considering not even all humans are hireable, I’d say only a fool aims for a simulacrum.
Well, it’s established wisdom that the dataset size needs to scale with the number of model parameters. Quadratically, IIRC. If you don’t have that much data the training basically won’t work; it will overfit or just not progress.
You also have to filter out the AI generated garbage that is rapidly becoming a majority of content on the internet.
…and even AI-generated “synthetic data” as options.
HAHAHA
This is how the best chess and go computers got to be as good as they are. AI generated “synthetic data.”
Yes and no.
Chess bots (like Stockfish) are trained on game samples, with the goal of predicting what search path to keep looking at and which moves will result in a win. You get game samples by playing the game, so it made sense to have stockfish play itself, since the input was always still generated by the rules of chess.
If a classifier or predictive model creates it’s own data without tying it to the rules and methods in reality, they’re going to become increasingly divorced from reality. If I had to guess, that’s what the guy in the article is referencing when talking about “sanitizing” the data. Some problems, like chess, are really easy. Mimicking human speech? Probably not
Yeah, because the human developers know the rules of chess, so it’s easy to generate or verify perfect quality games at massive scale. Natural language can’t be tackled like that; certainly not yet, probably not ever. Many have tried and failed to parse natural language algorithmically, but at the end of the day it seems to rely heavily on loose conventions and endless shared experiences. So, you need content from the wild, or you’re basically letting the AI mark its own homework.
I work in AI. This is very common, and lots of companies use this. It’s also very common in academia, as it’s an easy way to get data. Synthetic data can range from totally fake to techniques like machine translation to transform data from one language to another.
When they say “AI generated”, it’s probably just using one of the API’s the LLM orchestrates.
The human centipede, but circular.
Your laughter is misplaced. Synthetic data is a serious solution, and when it’s done right it can give better results than raw “real” data alone.
In other news, the world’s wealthiest people are running out of money after burning through the entire planet. Sources say one of the world’s multi-billionaires purchased a law firm that was in bed with the RIAA roughly 10-15 years ago when music piracy was supposedly costing more money than the GDP of all the peoples of the world, combined. “The Owners” (as they have recently rebranded) have decided to collect on this unpaid debt from every living soul, and from all the multinational companies who have been long-established as having no living souls whatsoever. A nameless, faceless, pitiless representative was quoted as saying: “Resistance… is futile. Your life, as it has been, is over. From this time forward, you will service… us.”
Is it wrong that I hope it eats itself and implodes?
If it’s wrong, then I’m wrong right along with you.
You’re rooting for a revolutionary new technology to fail rather than get better. I’d call that wrong.
If nothing else, AI is never going to get worse than it is now. So if that’s intolerably bad for you then improvement is the only way out.
AI is never going to get worse than it is now
Is that just a wild assumption, or…? One phenomena that has already been witnessed with AI is that it does in fact get worse if it trains upon it’s own output.
Given that I have locally-run AIs sitting on my home computer that I have no plan to delete (until something better comes along), then yeah, it’s never going to get worse. If all else fails I can just use the existing AI for as long as I want. It doesn’t “wear out.”
It doesn’t “wear out.”
The physical components will, and compatible components for older systems keep getting harder to come across. Computers are not immortal entities. Maintenance of older machines will continually become more labour and cost intensive over time.
The models are digital, making copies for safekeeping is easy.
The hardware is a computer, and computers are general-purpose. The kind that run AI models well at infrastructure scale are rather high end, but are still available off-the-shelf.
Computers are general-purpose machines. You can run a computer program on any computer, it may just be faster or slower depending on the computer’s capabilities.
The AIs I run locally are also open-source, so if future computers lose compatibility with existing programs they can be recompiled for the new architecture.
I suppose we could lose the ability to build computers entirely, but that strikes me as a much bigger and more general issue than just this AI thing.
You can run a computer program on any computer
Incorrect. Certain programs require certain standards for how the hardware is designed. There are already lots of old programs which can’t be run natively on modern machines, and using software to emulate a compatible environment can impact performance in more ways than just speed.
You’re wildly wrong about the fundamentals of computer science here. I’d be starting from first principles trying to explain further. I recommend reading up on Turing machines, or perhaps getting ChatGPT to explain it to you.
You’re rooting for a revolutionary new technology to fail rather than get better
As long as the oligarchs who run and own these AI systems are at the helm, yes I’m rooting for it to fail. Better is in the eyes of the beholder. Because come on, we all know better is going to be defined as better for the oligarchs, not you or me.
I run my own AI models on my own home PC. Am I an oligarch?
There’s nothing ‘revolutionary’ about a mass theft machine until EVERYONE IT’S STEALING FROM is getting paid out of the thieves’ pockets for what was stolen from them; and the people that run it make no profit from it. Til then, it’s just business as usual out of the west’s necrocapitalists; and your business makes me vomit.
Why do you feel this way?
Look at my profile for one reason and look at this video for another: Jon Stewart On The False Promises of AI | The Daily Show
More? We’ve got training data at home
How is the entire internet not enough to have perfected this shit?
Easy – their methods aren’t sufficient to begin with. No amount of training data would be enough. But perhaps they can develop new methods with what they’ve learned.
deleted by creator
While the article makes a big deal about a lack of data and even hint at synthetic data as an option, the truth is synthetic data is already being used and is just as good apparently at training. Such a misinformation article designed to stir the AI haters especially the headline.
They seem to be experimenting with that for sure, but need to ensure quality of the model doesn’t degrade, as per source article:
Anthropic’s chief scientist, Jared Kaplan, said some types of synthetic data can be helpful. Anthropic said it used “data we generate internally” to inform its latest versions of its Claude models. OpenAI also is exploring synthetic data generation, the spokeswoman said.
Imo we’ve clearly hit a limit with vertical scaling of data. We need some kind of breakthrough on better ways to process what data we’ve got if we want to continue making meaningful progress.
So, basically, back to the way the field was for the preceding 60 years.
Source WSJ article without paywall:
https://archive.is/R06ay
