causal

That was pointed at Crowdstrike - the authors of the study - who should definitely have that skill level.

Dude - I can't believe we're at the point where we're publishing headlines based on someone's experience writing prompts with no deeper analysis whatsoever.

What are the exact prompts and sampling parameters?

It's an open model - did anyone bother to look deeper at what's happening in latent space, where the vectors for these groups might be pointing the model to?

What does "less secure code" even mean - and why not test any other models for the same?

"AI said a thing when prompted!" is such lazy reporting IMO. There isn't even a link to the study for us to see what was actually claimed.

I love this sort of self-starter experimenting. Curious what models have been tried, I saw Grok4 mentioned, curious how well it transfers to other models.

Source?

Yeah I always hesitate to use emojis in any document or design for this reason, you have no idea how it's going to look to other viewers

The article gives some examples of scope creep but missed the biggest one IMO: copyright enforcement. I suspect if you follow the money, copyright is what keeps things like Chat Control coming back. Fully expect Sony, Disney and other IP to be added to the list of flagged content, keeping us safe from dangerous pirates.

I think there's something uniquely distracting about the constant availability of phones. We have muscle-memory now that can reflexively open a little hit of reward anytime we're in an idle moment.

Now instead of choosing to open our phones, we have to actively choose NOT to let that muscle memory spring into the action of unlocking the phone. Seems bad.

I generally find rootless pretty easy, it's just annoying that it's an additional few steps. Feels like an afterthought when it should be the default.

And robs children of family and play time. Play is critical work for the child mind.

You're absolutely right about the HN bias, but we're also a crowd with a lot of experience in some of the jobs supposedly being replaced and with the tools supposedly replacing us, and it's a little hard to buy.

The data about new grads not being hired isn't lying, but if AI is to blame we should be seeing a flood of data and stories on how roles X, Y and Z were obliterated by AI. But I have yet to see a single solid example, so I find it hard to believe it's a massive shift.

If anything, I would expect AI to be replacing older, expensive workers with young, tech-savvy workers who can put the AI to use. But that is clearly not the case.

This article discusses attention in a very immediate sense, but I think most of the points also apply to long-term attention.

Our behaviors are determined by habit far more than anything, willpower is seldom enough to result in behavioral patterns over time. Even things like the career we chose become habit; pivoting from technology to horticulture will not happen if you cannot change your daily habits to go from thinking about technology to thinking about horticulture.

This did not go where I thought it was going, and I'm glad. I enjoyed the read. I'm not versed enough in psychiatry to validate the brain-chemistry stuff but my practical experience lines up.

Reminds me of the trick of telling yourself "let's give this my full attention for just 5 minutes, and if I still don't want to do it we can move on". I pretty much always end up wanting to keep doing that thing.

Also AI hype is sucking all capital out of traditional hiring

I also have yet to see anyone meaningfully differentiate between "AI is taking jobs" and "AI hype is causing stupid executive decisions" and even "AI hype is sucking up all the capital that would normally go towards hiring".

Then in a subscription economy- where having a poor credit score can result in loss of housing, insurance, and more- opaque unelected beureaus are wielding tremendous power. Perhaps they are not evil or even unfair, but if you'll recall the context of TFA, it bears many of the same failure modes of a social credit system, and will only be as good as the people regulating it.

Many examples were given, please read the linked article. And again, in a society where everything is monetized, you can argue that creditworthiness is a just proxy for anything. So if basic water and electric access or housing access can be, so can anything else where you pay money for a service. You could argue dating apps should have access to credit scores and I wouldn't be surprised if some tried.

> Which institutions?

Credit bureaus...

Again, missing the point - it is a system that has become a centralized digital scoring mechanism. This is dangerous because it centralizes power, giving institutions that are not even under democratic control the power to punish individuals or groups as they see fit across all sectors of life.

Hey I see you're quibbling on details, maybe because you feel determined to win an argument to save face or something. I must point out that nobody here cares whether you were wrong, and digging in your heels even after multiple people provided multiple examples just drags the conversation to unproductive places.

As I said in the first reply, there is very little behavior you cannot tie money to in some way, so you can go down this path forever of trying to find a reason why credit scores are totally valid proxies for all human behavior, and you would be technically correct while missing the point entirely.

And a digital social credit system can be manipulated to drive behavioral changes en-masse. It centralizes power. Harder to do that with word-of-mouth.

Credit scores are used for a lot more than finances. You could argue everything they're used for is tied to money some way, but since everything is being sold to us as a service that doesn't leave much out.

They are also very much not decentralized.

Just look at this comment section - one flawed headline is all it takes to get hundreds of people writing essays about how they totally understand how the brain works and knew it all along.

The irony. It isn't even a new study. Way too much has been written about this flawed study when we should just be doing more studies.

Instead of trying to balance dreck can we just... not upvote any dreck

Yeah I feel like HN is being Reddit-ified with the amount of reposted clickbait that keeps making the front page :(

This study in particular has made the rounds several times as you said. The study measures impact of 18 people using ChatGPT just four times over four months. I'm sorry but there is no way that is controlling for noise.

I'm sympathetic to the idea that overusing AI causes atrophy but this is just clickbait for a topic we love to hate.

Yeah I dislike that term, because it puts the responsibility on the AI instead of the humans behind it. You wouldn't call it "paper psychosis" if OpenAI had written a letter encouraging psychotic behavior.

Prompt-injected SWATing even.

"know" is used colloquially this way even for non-AI systems.

"it encodes the training data in weights to predict a token mimicking a human ..." - better?

I don't think we should have to choose between "sycophantic coddling" and "alert the authorities". Surely there's a middle ground where it should be able to point the user to help and then refuse to participate further.

Of course jailbreaking via things like roleplay might still be possible, but at the point I don't really blame the model if the user is engineering the outcome.

It is disturbing, but I think a human therapist would also have told him not to do that, and instead resorted to some other intervention. It is maybe an example of why having a partial therapist is worse than none: it had the training data to know a real therapist wouldn't encourage displaying nooses at home, but did not have the holistic humanity and embodiment needed to intervene appropriately.

Edit: I should add that the sycophantic "trust me only"-type responses resemble nothing like appropriate therapy, and are where OpenAI most likely holds responsibility for their model's influence.