[bpodgursky]

After seeing Alpha* solve Go, Chess, and protein folding in the past ~3 years, I think it would be pretty silly for your prior to be discounting any Google AI project as vaporware.

Their models accomplish ridiculously powerful things. Tbh I think it's far _more_ likely the answer is "this is crazy powerful, but the engineers didn't feel like writing a blog post about it, and the marketing team hasn't figured out how to monetize it yet".

[MathYouF]

If there's anything SoTA AI researchers love and have experience doing it's writing blog posts and papers explaining how.

The lack of details makes me think they're either hiding a new technique they'd rather keep secret because it provides a competitive advantage, or that it's really only a marginal improvement over existing NLP models (or an ensemble of them with nearly no improvement on any given metric) and the 1000x improvement is on a metric that no actual ML scientist would respect.

I don't have the slightest bit of information about Google's AI team to know if those are the only two options and if so which is more likely.

[Lyapunov_Lover]

It's not a secret at all. Transformer models scale. Big models are powerful. Everyone knows this. Google can afford to train very big models. It's not a new technique. I think the issue here is that people are uncomfortable with the idea of AI models displaying scale relativity.

[visarga]

Big model also means lots of data, including lots of unfiltered garbage used in training. Nobody can manually review so much data, all they can do is automated filtering at this scale. So this means the model has a large attack surface and it is going to be used to do something bad and shame itself when put together with critics determined to find those gaps.

We have seen in the last few months attacks on Google Translate, GPT-3 and other language models from the PC crowd, including the famous AI Ethics firings. It's just tricky to show it in this climate.

The PC crowd don't believe language is fair and concepts neutral, instead saying they are an expression of systems of power. So language models are a natural target for them because they could amplify biases against their identity groups.

I find this critique hasty especially because big language models are nascent technology. We shouldn't throw away the baby with the bath water!

[Lyapunov_Lover]

The PC crowd is right. Language encodes our cultural beliefs, and many of them are pretty rotten. But how do you update a culture's shared set of beliefs? Banning words is a symbolic exercise. What we tend to do instead is that we tell stories and share perspectives. We learn to empathize.

Figuring out how to feed language models with diverse sources of information is a tough challenge, but not impossible. I share Gebru's concern about "stochastic parrots".

[visarga]

I'll take logical reasoning over "stories" any day.

And calling language model "parrots" is flouting. Many people worked for decades to reach that accomplishment, here come the critics to shit all over it.

> But how do you update a culture's shared set of beliefs?

It's not the place of AI models to do activism, and it's a slippery slope leading to AI based inquisition. Take a look at how China uses AI to oppress their own people.

[Lyapunov_Lover]

Stories are compressed representations of complex spatiotemporal patterns. We use stories to make sense of the world and to share our insights with others. And if you think about it, stories are essentially containers for if-then relationships. So they're not as far removed from logical reasoning as you might imagine.

I don't understand why you'd find the use of the term 'parrot' offensive. Language models extract linguistic patterns. GPTs generate patterns based on those which they have been trained on. That's a process that can be described as parroting. If you find it offensive because you think it implies that the researchers coming up with these models aren't worthy of credit, I think you are reading something into it that isn't there. At least not from my perspective.

When I mentioned updating cultural beliefs, I was referring to the traditional way of going about it: through cultural products. My point was that the "PC crowd" would be better off if they relied on this strategy rather than attempting to halt the development of language models. I was absolutely not suggesting that language models should be used to "train" members of society. That's a dystopian nightmare.

[visarga]

I think showing the model would immediately trigger the critics to nitpick it like the famous "He is a doctor. She is a nurse." case, so they just don't show it until they figure out a way to avoid that. Moreover, language models are easy to trick into politically incorrect conversations and porn. AI Dungeon's GPT-3 was writing lots of porn, for example.