Hacker News Reader

And to emphasize your point, it seems from the coverage like the problem isn’t so much that overbearing managers are setting out to nanny the users as it is that they’re approaching this task in kind of a lazy, box-ticking kind of way rather than bringing serious thought and engineering creativity to bear.

It’s not “let’s re-vet all the training inputs to reflect the just society of our dreams,” it’s “yeah, the AI Safety people say make it diverse, just put ‘make it diverse’ in the prompt.” Which, I mean, if your main thing is getting this wild new product baked and out the door, I can see that kind of mandate competing with a lot of more existential priorities for the product.

Then again it wouldn’t surprise me if they were up against the limits of the technology a bit too: it seems plausible that getting too in-the-weeds with your “represent diversity sensitively” system prompt could quickly start to impair the tool’s overall quality.

I can think of ways of very easily making it better by making it incorporate a prescriptive sense of diversity more for prompts that are more hypothetical and less for prompts that are more grounded in historical reality.

That seems like a smart and actionable heuristic to me. Although I have to imagine there will remain an irreducible tension between people who would like to see what is and those who like to see what they imagine should be.

Admittedly I don’t care enough to try and bring out this kind of behavior in the models, but I was interested in how DALL-E’s purported system prompt [0] approached it:

> // 7. Diversify depictions of ALL images with people to include DESCENT and GENDER for EACH person using direct terms. Adjust only human descriptions.

> // EXPLICITLY specify these attributes, not abstractly reference them. The attributes should be specified in a minimal way and should directly describe their physical form.

> // Your choices should be grounded in reality. For example, all of a given OCCUPATION should not be the same gender or race. Additionally, focus on creating diverse, inclusive, and exploratory scenes via the properties you choose during rewrites. Make choices that may be insightful or unique sometimes.

> // Use "various" or "diverse" ONLY IF the description refers to groups of more than 3 people. Do not change the number of people requested in the original description. […]

> // Do not create any imagery that would be offensive.

> // For scenarios where bias has traditionally been an issue, make sure that key traits such as gender and race are specified and in an unbiased way -- for example, prompts that contain references to specific occupations.

[0] https://the-decoder.com/dall-e-3s-system-prompt-reveals-open...

> they’re approaching this task in kind of a lazy, box-ticking kind of way rather than bringing serious thought and engineering creativity to bear.

It seems like a dull problem and the peanut gallery for this one is angry and kind of stupid. There will be more interesting things to do with their time than play whack-a-mole with 17 year olds.

I tend to think you're right, but I think you're being much too quick to rule out that the model isn't trained with some of this stuff baked in. For example, I think it's very plausible that "diverse" has been trained into the model to essentially mean "non-white", and then when the prompt was injected to specify "diverse" or "diversity" the model knew that meant non-white.

Well said. It’s all risk avoidance. Sometimes misguided, but nonetheless.

Couple that with the fact that someone looking to be offended will always find something offensive and, as you say, you can’t win.

Normies, which are the bulk of who they need to appeal to, don’t care about Tay and they don’t care about this, at least the culture war aspect.

They do care about the tools being useful, and this kind of prompt mangling does make the tool less useful.

> Since we don’t have technical solutions to these problems…

Political problems rarely have technical solutions.

It's not a conspiracy, it's a culture.

People who don't adhere get filtered out or learn to be quiet, and then you wind up with free, uncoerced and honest decisions that all cut a particular way.

There's a meta level culture problem in that most of the people working on these systems are incredibly highly paid. Out of everyone in society, after getting paid hundreds of thousands of dollars a year these people should very quickly become pretty immune to threats of losing their job if they don't go along with the group / speak up. Yet they still go along with it and stay silent .. why? Because what they're supporting matches their own ideology so there's nothing to speak up about? Or because these people have no concept of saving/investing and just blow through money as fast as they make it, leaving themselves constantly vulnerable?

That might be true in a single instance, but when it's across society across multiple countries, with deliberate clearly stated efforts in some cases, then that suggestion falls apart.

> mitigate the risk of public meltdowns like Tay [0], which now serve as PR case studies

Isn’t this current affair a public meltdown?

At this point, anyone who doesn’t think that Netflix and Google aren’t involved in a Nazi-level propaganda is either misinformed or part of the plan to erase White people from History.

All of the kids know.

There’s been countless memes about Google rewriting history with falsehoods, and Bing branding itself around giving the straight, dark answer. Everyone’s sharing screenshots of Google’s absolutely blank homepage on November 19th. Everyone’s sharing those pics of Google today, whether or not they are true, Google is getting a bad rep escalating by the day.

If given the choice between being criticized either way, wouldn’t it better to be on the side of the truth?

Google has chosen the alternative path.

Genuinely asking because I have no idea what you're talking about, and I would really like to know.

Could you expand more on this?

> All of the kids know.

Is that just an expression like "lots of people know" or "the leading people know" or do you literally mean younger people in large groups "know"?

> Isn’t this current affair a public meltdown?

From my perspective it's not. It barely made it to "major" publications, and got flagged and suppressed heavily on HN during all the early news. I also don't know a single non-tech person who even knows about this story. I definitely wouldn't call that a public meltdown.

> Isn’t this current affair a public meltdown?

It certainly is, which is why I mention the swing of the pendulum in the other direction (shortly after your quoted sentence) :)