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

Sharing a ChatGPT Account with My Wife

startupbaniya.com

Post reply on HN
Print view View original

51–60 of 67 posts

Re: Sharing a ChatGPT Account with My Wife

#51

Post by mhh__ »

kbelder wrote: post #48
mhh__ wrote: post #26

A pet theory of mine is that women will be better at using AI for all but the most analytical/symbolic work (even this may not hold but I'd imagine the crossover is around there) I have no mechanism but it feels true.

>I have no mechanism but it feels true. I really, really want to make a joke here, but it would be so sexist I'd get voted down in an instant.

I'm being quite serious. I suppose the closest thing to an analytical description of what I have in my head is that the way I've seen women prompt chatgpt is probably much closer to how its trained as part of RLHF, but more deeply it seems like a very different way of thinking that men overall aren't particularly good at.

I met a sculptor recently (reasonable thirst for knowledge; but not from greece) who went straight for deepseek to ask a question. That surprised me. Obviously deepseek was in the news, but the way they train seems to be more aligned with the aforementioned way of thinking than openai.

Re: Sharing a ChatGPT Account with My Wife

#52

Post by makeitdouble »

scarface_74 wrote: post #44
viccis wrote: post #4

>We used to struggle with financial decisions—where to invest, whether real estate was a good idea, when is the right time to buy a house, pension plans, S&P 500, Bitcoin, all of it. We never really figured it out properly. Now, we just ask AI, and it helps us structure our thinking. This one is a step too far for me. Glad it's working out for you (for now) though.

He said it “structures their thinking”. Think in terms of a Rogerian psychotherapist that just keeps asking you questions and makes you think about things you wouldn’t have thought about. He didn’t say he would use the advice. I use ChatGPT all of the time to clarify my thoughts like this and NotebookLM actually comes up with some good questions I didn’t think about when doing discovery with a client (cloud consultin…

The scarry part is before using ChatGPT they couldn't figure it out, so they're either getting more information or are ending with the same results but just feel better about it, whithout still having figured it out.

It sounds innocent enough, but your parallel to psychotherapy is I think very spot on. You need trust in your therapist and they also do their best to be non directive and let you explore your own state. When you tell your therapist you have self-harm thought, it makes a world of difference when they ask what happened vs what tools do you think you need.

To me, my train of thought is the last part I want involvement from a third party I have no trust in.

Re: Sharing a ChatGPT Account with My Wife

#53

Post by makeitdouble »

viccis wrote: post #4

>We used to struggle with financial decisions—where to invest, whether real estate was a good idea, when is the right time to buy a house, pension plans, S&P 500, Bitcoin, all of it. We never really figured it out properly. Now, we just ask AI, and it helps us structure our thinking. This one is a step too far for me. Glad it's working out for you (for now) though.

> My wife used to draft simple business contracts by mixing and matching content she found in the public domain. Now, she uses AI—better quality, less time spent.

That part sounded way worse to me. In the end she may be fully understandind the contracts and ramifications and it doesn't matter where the bits and parts came from...but that really doesn't sound like it, and we're talking legal documents.

Re: Sharing a ChatGPT Account with My Wife

#54

Post by rat87 »

viccis wrote: post #4

>We used to struggle with financial decisions—where to invest, whether real estate was a good idea, when is the right time to buy a house, pension plans, S&P 500, Bitcoin, all of it. We never really figured it out properly. Now, we just ask AI, and it helps us structure our thinking. This one is a step too far for me. Glad it's working out for you (for now) though.

Just put it in an index fund.

Re: Sharing a ChatGPT Account with My Wife

#55

Post by makeitdouble »

elliotto wrote: post #45
voidhorse wrote: post #5

This post is a good piece of evidence that the value of LLMs is driven far more by the circumstances surrounding them than by positive value they bring. All I see is a society in which people are stretched insanely thin (even the software engineers work on monetized side projects, for god's sake), where basic needs like proper education and time to focus are degrading or too expensive, where information gathering has…

I think almost all N are a suboptimal solution to what are inherently structural social problems that arise from a distribution of wealth and organization of labor where N is any technological invention of the past 20 years. Open source insulin lol the empire is collapsing.

You are right that technology isn't a cure for solving societal problems. Now some societies getting worse and worse in the last decades doesn't invalidate the better uses of technical advances, at least they have tremendous impact in countries that care more about their citizens.

Re: Sharing a ChatGPT Account with My Wife

#56

Post by rat87 »

viccis wrote: post #4

>We used to struggle with financial decisions—where to invest, whether real estate was a good idea, when is the right time to buy a house, pension plans, S&P 500, Bitcoin, all of it. We never really figured it out properly. Now, we just ask AI, and it helps us structure our thinking. This one is a step too far for me. Glad it's working out for you (for now) though.

"Just put it in an index fund"advice has been around far longer then llms, as well as "time in the market beats timing the market"

Re: Sharing a ChatGPT Account with My Wife

#57

Post by lolinder »

glenstein wrote: post #47
everdrive wrote: post #39

Earlier quoted context omitted.

It doesn't even have to be ads. The people who run it have their own ideological sense of right and wrong. And no matter what you think of their moral stance so far, there's nothing to prevent them from changing their values in the future. If you rely on them to think, you're also letting them cordon off certain kinds of thoughts.

And this is where it gets into difficult questions around objectivity and misinformation. Because objective knowledge is more under threat now than it has ever been. And it's in part because we don't have an agreed upon consensus canon of knowledge. Or rather, there's at least skepticism of such things. But I think there's good reasons for believing that people as individuals and through collective efforts at institu…

You're conflating a few concepts:

* Objective knowledge

* Consensus knowledge

Consensus knowledge is not the same thing as objective knowledge, nor is it in and of itself a Good Thing.

Consensus knowledge is simply the current consensus of a specific group of people, who themselves bring a ton of subjectivity to their analysis. The smaller that group and the more self-selecting it is, the lower the value of their consensus, because of necessity they represent only a tiny fraction of the sum total of human experience. And the larger the group is and the more dynamic the selection process, the more difficult it becomes to summarize their perspective into any sort of consensus knowledge.

Objective knowledge is not consensus knowledge, objective knowledge is the platonic ideal to which consensus knowledge aspires.

Conflating consensus knowledge with objective knowledge is how we ended up in a place where so many people across the board question the idea that objective truth even exists! Instead of where the scientific method started—with a philosophy built around the idea that there is objective truth and we are capable of better and better modeling it through rigorous tests—we're in a place where some people have conflated the map for the territory, leading others to question whether the territory exists at all.

Re: Sharing a ChatGPT Account with My Wife

#58

Post by glenstein »

lolinder wrote: post #57
glenstein wrote: post #47

Earlier quoted context omitted.

And this is where it gets into difficult questions around objectivity and misinformation. Because objective knowledge is more under threat now than it has ever been. And it's in part because we don't have an agreed upon consensus canon of knowledge. Or rather, there's at least skepticism of such things. But I think there's good reasons for believing that people as individuals and through collective efforts at institu…

You're conflating a few concepts: * Objective knowledge * Consensus knowledge Consensus knowledge is not the same thing as objective knowledge, nor is it in and of itself a Good Thing. Consensus knowledge is simply the current consensus of a specific group of people, who themselves bring a ton of subjectivity to their analysis. The smaller that group and the more self-selecting it is, the lower the value of their con…

I'm not sure I understand where you're seeing a conflation? I promise you that I was already familiar with the notions of objective and consensus as you described them.

I acknowledge that these are distinct concepts, but there's nothing in what I said above where I'm equating them. At least on charitable interpretation.

I'm also not sure what I agree with much or any of your supplementary analysis. Just to pick one example, I don't think our confidence in the measurement of the Higgs Boson is in any way contaminated by the fact that the consensus on it exists within a self-selected academic community, or that it fails to sufficiently sample the global population.

Re: Sharing a ChatGPT Account with My Wife

#59

Post by makeitdouble »

rat87 wrote: post #54
viccis wrote: post #4

>We used to struggle with financial decisions—where to invest, whether real estate was a good idea, when is the right time to buy a house, pension plans, S&P 500, Bitcoin, all of it. We never really figured it out properly. Now, we just ask AI, and it helps us structure our thinking. This one is a step too far for me. Glad it's working out for you (for now) though.

Just put it in an index fund.

The interesting part of this advice: index funds will probably include Starlink, Tesla, Apple, Amazon, Meta and Microsoft etc.

While "index fund" sounds plain and boring, it's an inplicit vote for the status quo and binding one's fortune to the current market leaders, effectively supporting them.

Re: Sharing a ChatGPT Account with My Wife

#60

Post by jcims »

sumedh wrote: post #15
viccis wrote: post #4

>We used to struggle with financial decisions—where to invest, whether real estate was a good idea, when is the right time to buy a house, pension plans, S&P 500, Bitcoin, all of it. We never really figured it out properly. Now, we just ask AI, and it helps us structure our thinking. This one is a step too far for me. Glad it's working out for you (for now) though.

> when is the right time to buy a house, pension plans, S&P 500, Bitcoin, all of it. This is why I think ChatGpt can become the world's biggest ad company. Any recommendation it gives can be go through a bidding network where advertiser can bid for their keywords and ChatGpt can give the highest bidders recommendation. Of Course it needs to give good or good enough recommendation, if the quality goes down then users…

I built a custom gpt to inject 'covert' advertising into its responses with varying levels of transparency. At more covert levels it wouldn't disclose any brands unless you asked it for specific examples of products that had the benefits it would talk about. At the most covert it wouldn't even do that, but would litter its responses with enough keywords that a google search would lead you to the product it was intending to pitch.

The wild part is that it would always lie about why it was doing it, not disclosing anything about the framework or intent of injecting that content into its responses. Yes at some level it's obvious it would do that, but it's also interesting to see it first hand.

Post reply on HN