Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Imagen, a text-to-image diffusion model

gweb-research-imagen.appspot.com

Post reply on HN
Print view View original

601–610 of 661 posts

Re: Imagen, a text-to-image diffusion model

#601

Post by ceejayoz »

nearbuy wrote: post #499
ChadNauseam wrote: post #233

Earlier quoted context omitted.

You know, it wouldn't surprise me if people talking about how black curly hair shouldn't be seen as unprofessional contributed to google thinking there's an association between the concepts of "unprofessional hair" and "black curly hair"

That's exactly what's happening. Doing the search from the article of "unprofessional hair for work" brings up images with headlines like "It's ridiculous to say that black women's hair is unprofessional". (In addition to now bringing up images from that article itself and other similar articles comparing Google Images searches.)

You’re getting cause and effect backwards. The coverage of this changed the results, as did Google’s ensuing interventions.

Re: Imagen, a text-to-image diffusion model

#603

Post by orblivion »

isx726552 wrote: post #599
beeskneecaps wrote: post #596

It’s terrifying that all of these models are one colab notebook away from unleashing unlimited, disastrous imagery on the internet. At least some companies are starting to realize this and are not releasing the source code. However they always manage to write a scientific paper and blog post detailing the exact process to create the model, so it will eventually be recreated by a third party. Meanwhile, Nvidia sees no…

I am absolutely terrified of all this for a different reason: all human professions (not just art) will soon be replaced by “good enough” AI, creating a world flooded with auto-generated junk and billions of people trapped permanently in slums, because you can’t compete with free, and no one can earn a living any longer. It’s an old fear for sure but it seems to be getting closer and closer every day, and yet most of…

And then once you take the (probably trivial) step where the computers come up with the ideas for the images, these images won't be interesting anymore because we know a human didn't even make it. It won't be funny in the same way. "Oh that was clever" doesn't make sense anymore. We could reach a new level of jaded.

(Also, hello readers from the year 2032 when all of these predictions sound silly.)

Re: Imagen, a text-to-image diffusion model

#604

Post by SemanticStrengh »

ccbccccbbcccbb wrote: post #247
nullc wrote: post #244

Earlier quoted context omitted.

Get offline and talk to people in meat-space. You're likely to find them to be much more reasonable. :)

Yep, the meat-space is generally a bit less woke than HN, so thanks for the reminder ))

Smoking these meats! https://youtu.be/YeemJlrNx2Q

Re: Imagen, a text-to-image diffusion model

#605

Post by kurthr »

orblivion wrote: post #603
isx726552 wrote: post #599

Earlier quoted context omitted.

I am absolutely terrified of all this for a different reason: all human professions (not just art) will soon be replaced by “good enough” AI, creating a world flooded with auto-generated junk and billions of people trapped permanently in slums, because you can’t compete with free, and no one can earn a living any longer. It’s an old fear for sure but it seems to be getting closer and closer every day, and yet most of…

And then once you take the (probably trivial) step where the computers come up with the ideas for the images, these images won't be interesting anymore because we know a human didn't even make it. It won't be funny in the same way. "Oh that was clever" doesn't make sense anymore. We could reach a new level of jaded. (Also, hello readers from the year 2032 when all of these predictions sound silly.)

Don't forget the training data for those computer "ideas" will be "attention" and targeted at the most vulnerable 80% of the market. I'd hope that it makes them less fearful and angry, but nope... that drives attention. I wonder what combination of UFOs, satanic cults, and immigrant hoards it will be.

Re: Imagen, a text-to-image diffusion model

#606

Post by SemanticStrengh »

nomel wrote: post #115
riffraff wrote: post #77

Earlier quoted context omitted.

This seems bullshit to me, considering Google translate and google images encode the same biases and stereotypes, and are widely available.

Aren't those old systems?

Pre woke tools, wouldn't have been allowed nowadays.

Re: Imagen, a text-to-image diffusion model

#607

Post by kvetching »

beeskneecaps wrote: post #596

It’s terrifying that all of these models are one colab notebook away from unleashing unlimited, disastrous imagery on the internet. At least some companies are starting to realize this and are not releasing the source code. However they always manage to write a scientific paper and blog post detailing the exact process to create the model, so it will eventually be recreated by a third party. Meanwhile, Nvidia sees no…

I, for One, Welcome Our Robot Overlords

Re: Imagen, a text-to-image diffusion model

#608

Post by kvetching »

discmonkey wrote: post #421

For people complaining that they can't play with the model... I work at Google and I also can't play with the model :'(

Good thing there is a company committed to Open Sourcing these sorts of AI models.

Oh wait.

Google: "it's too dangerous to release to the public"

OpenAI: "we are committed to open source AGI but this model is too dangerous to release to the public"

Re: Imagen, a text-to-image diffusion model

#609

Post by themeiguoren »

londons_explore wrote: post #557
mistrial9 wrote: post #119

Reading a relatively-recent Machine Learning paper from some elite source, and after multiple repititions of bragging and puffery, in the middle of the paper, the charts show that they had beaten the score of a high-ranking algorithm in their specific domain, moving the best consistant result from 86% accuracy to 88% accuracy, somewhere around there. My response was: they got a lot of attention within their world by…

If you worked in a hospital and you managed to increase the survival rate from 86% to 88%, you too would be a hero. Sure, it's only 2%, but if it's on a problem where everyone else has been trying to make that improvement for a long time, and that improvement means big economic or social gains, then it's worth it.

I like focusing on the failure rate instead - going from 14% to 12% is a pretty big jump.

Re: Imagen, a text-to-image diffusion model

#610

Post by visarga »

godelski wrote: post #72
visarga wrote: post #24

Earlier quoted context omitted.

The big labs have become very sensitive with large model releases. It's too easy to make them generate bad PR, to the point of not releasing almost any of them. Flamingo was also a pretty great vison-language model that wasn't released, not even in a demo. PaLM is supposedly better than GPT-3 but closed off. It will probably take a year for open source models to appear.

That's because we're still bad about long-tailed data and that people outside the research don't realize that we're first prioritizing realistic images before we deal with long-tailed data (which is going to be the more generic form of bias). To be honest, it is a bit silly to focus on long-tailed data when results aren't great. That's why we see the constant pattern of getting good on a dataset and then focusing on…

> some community members focus on finding these holes but not fixing them

That's what bothered me the most in Timnit's crusade. Throw the baby with the bath water!

Post reply on HN