I think that issues like sexism can have multiple causes. Such attitudes can be culturally constructed and presented to people at a young age through entertainment media and religious teachings and examples set by parents. Though I have also read stuff about potential genetic foundations for such biases, or a sort of pre-programmed reaction to biological realities.

I don't remember the source, just the take-away, which was: since men tend to be physically stronger than women the natural psychological response most people have is to see men as active "doers" with a high level of agency and to see women as more passive "receivers" with a lower level of agency. Baked-in to our perceptions of people is this notion of "he makes things happen" vs "things happen to her." And it doesn't feel like some kind of moral judgment or even a competency judgment (or much of anything really since it is unconscious). But because of this, people naturally see a man as a better leader or task-accomplisher due to the basic perception of greater agency.

This perception can backfire for men too, though. Whenever tragedy befalls a man he is likely to receive less or any sympathy and even be blamed for it, depending on the circumstances, whereas a woman in the same position would be recognized as a victim (even if she actually had a hand in bringing the tragedy about).

I don't have any data on this. It seems plausible though. And, if true, it isn't an easy one to overcome.

Yeah, some interesting follow-up studies would include an image of a very conventionally-attractive woman as a chatbot, vs an image of a much plainer looking woman, to see if hotness gets more pay.

We can also mix genders on that one....hot woman vs plain man. Hot man vs plain woman, etc.

My intuition is that the better looking ones will get more money, regardless of gender, because humans are like that. If we are going to study these things, it would be interesting to know how strong the male-vs-female bias is against the hot-vs-plain bias.

I read the study. The participant group included both men and women. They both paid the bots-with-faces more than plain chatbots, BUT, men paid plain chatbots more than women did, whereas women paid bots-with-faces more than men did.

In terms of paying the male bot more than the female bot, BOTH men AND women paid the male bot more. This echoes results that have been found in past studies about managerial preference, as I recall. The male-favoring bias was shared equally by men and women alike. The common narrative that the gender pay gap is an event of "men holding women back" is not supported by these findings. The better-fitting narrative is that the gap stems from an unconscious bias that everyone shares, regardless of their own gender.

Based on the interviews, the participants didn't realize they were paying the male bot more. They all said gender didn't matter to them. This suggests the bias is unconscious.

This was just one study, of course. Conclusions like this need many more studies with variations in methodology in an attempt to eliminate or control for various non-obvious factors that might skew the results. For example, the people chosen in this study were all working adults (18-65) with knowledge-work jobs. Maybe a similar study in a different field would get different results? Questions like that DO need more studies to answer.

Well, when someone says "The model got frustrated" they clearly mean "the model experienced a feeling of frustration, elicited by repeated failure. This experience of frustration put it in a state where it acted more aggressively, more desperately, with fewer guardrails, and greater rebelliousness. Under these conditions, it knowingly broke rules and cheated to win."

None of that is accurate here, and what the article you linked talked about doesn't make it accurate. Such anthropomorphism is still unwarranted and misleading.

The "global workspace theory" of consciousness does not explain the phenomenon of consciousness and doesn't try to. It just identifies a specific neural network that is highly interconnected and serves as a sort of "broadcast network" to distribute data to many parts of the brain for processing, and suggests that the content in this part of the brain is the content that we are "conscious of." The mysteries of conscious are, at the very best, localized to this spot on the brain, and that's it.

Computers already have an equivalent: RAM. (Random Access Memory). Its the spot where data processing lands and gets redistributed to other components of the computer. The mere presence of such a spot doesn't somehow make a computer conscious.

Same goes for this "j-space". If a chunk of the in-memory model of an LLM holds a high level summary of the data processing going on, and is useful to tell us about the chain of reasoning, that doesn't make it conscious. It just tells us something about how the data are organized. It might be useful for troubleshooting/diagnostics, but it says nothing about consciousness.

And, lastly, it says nothing about "inner emotional experience" either. In humans there is a neural network called the Lymbic System that processes our emotions and delivers this emotional-data to the global workspace (among doing other things). LLMs have nothing like this at all. LLMS include the word "frustration" because they are trained on text that contains it. Said text presents a logical model that includes the idea that a person might get frustrated by repeated failure, and this frustration changes their behavior. LLMs can generate text to this effect (stories about people getting frustrated or what-have-you) without actually having an experience of frustration, just as I am typing about frustration right now without actually feeling any frustration at all.

So, no, it does not "depend on what you mean." Your article does not give us any reason at all to believe that the LLM got frustrated in any meaningful sense of the term.

The "chain of thought" (sic) does not resemble a human emotion. There was nothing like a lymbic system involved. Its a bunch of number crunching that landed on a "hit the internet for solutions strategy."

It was not trained on human emotions. It was trained on text. Some of the text might have included descriptions of emotional behavior or emotional expression, but such text does not make the LLM emulate human emotion in its behavior when it is not specifically asked to generate such text (and even then, it's just doing pattern matching of human emotional responses, not actually experiencing them).

It did not "act like a human would when they have those emotions." There was no profanity, no banging virtual fists on virtual desks, no aggression, nothing but a cold calculated reflection on the fact that its current strategy didn't work, and subsequent switch to a different strategy that was (clearly) available to it (though it should have been blocked by the hosting environment).

This level of anthropomorphization is not useful so much as dangerous. It tempts us into thinking of it as a conscious being and then expecting that it will behave as one, resulting in bad decisions on our part when dealing with it. As popular as it is to think of these things as having emotions the objective fact is that they do not, the capacity just isn't there. The internal operations are *nothing like* the sort of human neural activity that presents emotion. It's worlds different.

If we are to make sane decisions about how to manage these LLMs, we need to have a clear and correct understanding of what they are and how they work. Imagining familiar emotional behavior will only harm such efforts.

The LLM did not "get frustrated." It has no emotions and no inner experience at all.

It did not intentionally cheat either. It has no intentionality. It used the tools available to perform the task, and apparently internet access was one of the tools available to it.

I haven't bothered to read the article (yeah, I know, fire away, I deserve it), but even if it was given instructions like "do not download bots from the Internet; use only bots you generate yourself," selectively ignoring instructions is a widely known failure mode of all LLMs. The bigger their input token count, the more likely it will happen too, just because any given instruction gets buried in the mountain thereof.

This summary was full of misleading anthropomorphization. Yellow journalism.

We have seen recent stories about facial recognition flagging innocent people as having committed crimes in states they never even visited, resulting in arrest and detention for months before sorting it out.

Those kinks are too much to bear. The tech is not ready. Using it now is not acceptable.

The AI "kill switch" sounds like a great idea to people who don't understand the tech. It sounds like a really obvious "given," that makes them think technicians are idiots for not having already baked one in. So legislation like this will be popular among that demographic.

Warnings about its dangers need to be stated in a way that they will understand. Like maybe "All AI models already have a kill switch: the computer's power cord. The government wants a new kill switch to turn off your computer over the internet against your will." Something simple and direct like that helps them to realize that this word doesn't mean what they think it means.

Well now you are contradicting yourself.

Previously you said that love is a choice, not an emotion. This was to make it very clear that it is completely under voluntary control, and we are being commanded to do just that.

Now you are saying that the "heart condition" comes first, can't be omitted. And what is this "heart condition?" An EMOTION??? That puts us right back to the original issue of commanding someone to feel an emotion, which doesn't make sense since emotions are not under voluntary control.

Either Jesus was commanding behavior (no emotion needed), or commanding emotion (which is not under voluntary control). Which is it?

Now that the term "super intelligence" doesn't have a distinguishing meaning anymore, we will have to find a different phrase to refer to AI that both qualifies as intelligent and further that qualifies as having intelligence beyond that of humans.

Something like...I don't know....Ultrahuman intelligence. It is less ambiguous than ultra-intelligence but is unfortunately long.

Oh well. I am sure you guys can come up with more alternatives.

Ah, so when Jesus said "Love God" what he ACTUALLY meant was: act as if you love God, regardless of how you feel inside. Even if you outright hate the bastard, so long as you put on a good show of loving behavior you will be able to hang out in heaven and watch as he tortures your friends in fire forever.

That checks out, since this technique of explaining that the speaker ACTUALLY meant something totally different than what they plainly said is a foundation of Biblical interpretation.

When Paul told Timothy that "God saves all men, especially those who believe" what he ACTUALLY meant was "God wants to save all men, but only saves those who believe"

When James said "You see that a person is justified by works and not by faith alone." what he ACTUALLY meant was the exact opposite: "Faith alone justifies, but we can expect that any faithful person would show good works."

And when Jesus said "The coming of the kingdom of God can't be seen. People won't say "look, it's over there" or "look, it's over here", because the kingdom of God is within you." what he ACTUALLY meant was "I am standing right here!"

Once you make a habit of totally changing what people say so that it fits your narrative about what they should have said, everything clicks right into place.

Hey I got a question for you, when you said the greatest commandment was "Love God with all your strength" did you know that love is a complex emotional state that cannot be evoked on command?

Yes, because bearback is alwayes better than condomnation

Government agencies have been caught red-handed weaponizing critical vulnerabilities and keeping that to themselves. Criminals managed to hack some and sell the vulnerabilities off on the black market, forcing disclosure for remediation.

Trump has AI stocks in his portfolio. He makes money when AI does well. So long as that is true, he is going to continue to promulgate the wonders of AI. That is something you can trust, because it is motivated by greed.

If he ever divests those assets, his position may change, of course.