This is among the least useful polls I have ever seen. Especially notable though, Kenya: 125 respondents = 12.0% of the survey, versus about 0.71% of world population India: 190 respondents =...
This is among the least useful polls I have ever seen. Especially notable though,
Kenya: 125 respondents = 12.0% of the survey, versus about 0.71% of world population
India: 190 respondents = 18.25% of the survey, versus about 17.79% of world population
Why am I interested in these in particular? Because of who this poll is really measuring.
The survey was conducted in seven languages across 64 countries, with participants recruited through Prolific, a paid research panel platform, ensuring inclusion of people across a range of income levels and contexts including those who do not use AI tools. The interactive format combined standard poll questions, open-ended prompts, and peer voting on each other’s statements.
I am deeply familiar with this subset of the population, as is anyone who has worked with Mechanical Turk, Upwork, and other platforms that offer mass scale “human” piecework employment. These people are among the heaviest adopters of AI tooling, as they automate their own processes to maximize their task throughput. I’ve spent a lot of time on the worker side forums reading about which micro jobs are best suited to automation, trading greasemonkey scripts to fill forms, documenting ban rates by employer, and so on.
All this to say: I would be shocked if AI hadn’t had its hands in directly answering a large percentage of these answers.
Full results here 72% saying AI made daily life better. People only trust their doctor more to act in their best interest. The trust ChatGPT or similar over social media, elected officials,...
People only trust their doctor more to act in their best interest. The trust ChatGPT or similar over social media, elected officials, community leaders, faith leaders, and civil servants.
Considering how they're algorithms designed to give satisfying-sounding answers at its core, even if you have stuff like guardrails and guidelines etc, I don't think it's that strange if you're...
Considering how they're algorithms designed to give satisfying-sounding answers at its core, even if you have stuff like guardrails and guidelines etc, I don't think it's that strange if you're not familiar with how they work.
In a sense producing a sense of trust is in their design, regardless or whether that's justified or not.
I totally expect that result. GPT and other LLMs are basically yes men that will speak in a reaffirming tone. Most people like that and build trust based upon it. I remember feeling some kind of...
I totally expect that result. GPT and other LLMs are basically yes men that will speak in a reaffirming tone. Most people like that and build trust based upon it.
I remember feeling some kind of connection when I first used an LLM as well. Because at the time I didn't really understand how they worked, I was not mentally prepared for how much self-affirming language it would use. Nobody ever spoke to me like that, so it genuinely felt good (even if the output was telling me I was wrong, it told me in just about the nicest, most reassuring way I could imagine). I think that applies to a lot of people, we're so used to not having others reaffirm us that even a chat bot "love bombing" us ends up filling that void, for some more than others.
And that's the dangerous part, since chat bots don't know what they're saying, they're just regurgitating statistically probable answers, which puts many people in vulnerable positions now that all they listen to is their output.
I eventually I learned how they operate and now try to avoid them like the plague, but most people don't leave that first enamored phase, they simply keep using LLMs to reaffirm whatever they're already thinking and end up trusting the output a lot more as a result.
There's another strange thing happening with LLMs and how people don't have their guard up about them. We are good at detecting things that are a little off and artificial. If these were robots...
There's another strange thing happening with LLMs and how people don't have their guard up about them.
We are good at detecting things that are a little off and artificial. If these were robots with almost human faces, and they talked to you and looked at you but they just had soulless eyes or moved a little strangely or their skin looked a little plastic... And if they praised you too much and sometimes said really weird or wrong things... We wouldn't trust them at all and we'd be creeped out.
But it's just text so we are not properly suspicious. We haven't evolved enough yet to see threats that don't have a face I guess.
Well, let's break down the list, because it doesn't actually seem that foolish after really thinking about it. Heavily dependant on the social media of course, but in general, I think I probably...
Well, let's break down the list, because it doesn't actually seem that foolish after really thinking about it.
social media
Heavily dependant on the social media of course, but in general, I think I probably trust things I read less on social media than just random chance. It's virtually all ragebait and nonsense, so this one makes sense to me.
elected officials
In the US, on a federal level, I think most of us can agree that most elected officials are literally professional con men at this moment in time. I tend to automatically assume whatever a federal official says is the exact opposite of the truth at this point and I'm usually proven correct. State and local vary by municipality, but in my case, it's mostly true for them as well.
community leaders, faith leaders
This one is honestly surprising to me. I'm not religious or involved with community groups, but why would you be involved in those organizations if you didn't trust their leaders? They're voluntary for the most part
and civil servants
Highly dependent on most civil servants. For the most part, they're informed and professional, but I have had experiences where it was clear that they just wanted me to go away, and they were intentionally misleading me to get me lost in the gigantic cogs of bureaucracy. It doesn't happen to me often, because I'm basically what society assumes is a "default human being", a middle aged married straight cis white male with a steady job, no major health issues, and no criminal record. The further you deviate from that "default", the more problems I'd expect you'd have with civil servants.
Even though LLMs are going to hallucinate and affirm your stupid ideas, depending on circumstances, I'd trust its answer more than a lot of the people listed above. Doesn't seem that foolish to me overall.
I interpreted this from a slightly different angle. Think of "faith leader" not as your minister or pastor, but as "the ministers in my community, who are ostensibly trying to make thier...
This one is honestly surprising to me. I'm not religious or involved with community groups, but why would you be involved in those organizations if you didn't trust their leaders?
I interpreted this from a slightly different angle. Think of "faith leader" not as your minister or pastor, but as "the ministers in my community, who are ostensibly trying to make thier communities better, in addition to preaching thier faith".
Even if the majority of people in a community "go to church", they don't all go to the same church, or even belong to the same faith. So it would be incorrect to assume that if someone goes to church and does trust their own minister, that doesn't necessarily mean they trust all of them.
The question was asking about trust in “your community or faith leader” so it is asking about the leader of your church specifically not just general community leaders
The question was asking about trust in “your community or faith leader” so it is asking about the leader of your church specifically not just general community leaders
I (kind of) trust it more over all of those people, but only because I can repeatedly push back against it and direct it as more of a research partner than an authority, and I don't trust those...
I (kind of) trust it more over all of those people, but only because I can repeatedly push back against it and direct it as more of a research partner than an authority, and I don't trust those people due to agendas, ignorance, laziness, or all three.
But I don't think most people actually push back and every time they show me their prompts, they aren't even asking the right question you would want to ask a real person, so it's a bit terrifying to me. But that's basically my takeaway anytime I encounter the thought processes of most people, so idk ¯\_(ツ)_/¯
I wonder what people are thinking of when they say AI. I think I'd agree with the statement that AI made daily life better, but I'd be talking about the ML algorithms that make modern search...
I wonder what people are thinking of when they say AI. I think I'd agree with the statement that AI made daily life better, but I'd be talking about the ML algorithms that make modern search engines work well, the ability for my smart cameras to detect people instead of random shadows, my car's lane keeping assist functionality, a lot of the ML functions software I use at work uses to detect anomolies and so forth.
LLMs are largely negative, although I do find them useful sometimes. That doesn't really outweigh their content polluting search results though. They still have a pretty minor impact on my life other than supply chain shortages though.
If I take the sum of "AI" as a whole though, yeah, mostly positive.
I think one of the negatives of the term AI has always been its fuzzy definition which constantly changes in colloquial usage. I strongly suspect that people are only thinking about LLMs and diffusion based image generation models when they hear AI.
This is interesting for more than AI insights. It also looks into how people want the future shaped. Ie People prefer guaranteed work over guaranteed income. It asks who people trust, with...
This is interesting for more than AI insights. It also looks into how people want the future shaped.
Ie
People prefer guaranteed work over guaranteed income.
It asks who people trust, with government right at the bottom.
Ah, OK, I was curious about that! I only made it ~a third through the survey, but the point about people desiring to work vs. having freedom to do what they please, after their jobs have been...
There is also a dignity factor. Many would prefer to work for their income rather than take handouts. Personally I'd vote for income over work. It means that those who can't work are still taken...
There is also a dignity factor. Many would prefer to work for their income rather than take handouts.
Personally I'd vote for income over work. It means that those who can't work are still taken care of. It also means that I'm not reliant on a workplace for survival. I'd hate to think what these guaranteed jobs are going to be like.
It sounds great to be able to do anything anytime. But it sounds pretty awful to have everyone else doing whatever they want at the same time as me! Plus, with no limitations, freedom can become...
It sounds great to be able to do anything anytime. But it sounds pretty awful to have everyone else doing whatever they want at the same time as me! Plus, with no limitations, freedom can become its own prison. It becomes the hedonic treadmill.
Would you mind if I ask about why that is? I'm not sure if it's digging into anything personal -- so please feel free to ignore if that's the case -- but that sounds fantastic to me, imo. I love...
But it sounds pretty awful to have everyone else doing whatever they want at the same time as me!
Would you mind if I ask about why that is? I'm not sure if it's digging into anything personal -- so please feel free to ignore if that's the case -- but that sounds fantastic to me, imo. I love weird stuff like gigantic, steel spider sculptures appearing in underpasses (despite having a touch of arachnophobia), or music festivals popping up in random fields, so people being free to pursue their passions sounds lovely imo.
Plus, with no limitations, freedom can become its own prison. It becomes the hedonic treadmill.
Q: aren't we basically talking about retirement, only early enough that peoples' bodies still function (and their friends are all still alive)?
This is among the least useful polls I have ever seen. Especially notable though,
Kenya: 125 respondents = 12.0% of the survey, versus about 0.71% of world population
India: 190 respondents = 18.25% of the survey, versus about 17.79% of world population
Why am I interested in these in particular? Because of who this poll is really measuring.
I am deeply familiar with this subset of the population, as is anyone who has worked with Mechanical Turk, Upwork, and other platforms that offer mass scale “human” piecework employment. These people are among the heaviest adopters of AI tooling, as they automate their own processes to maximize their task throughput. I’ve spent a lot of time on the worker side forums reading about which micro jobs are best suited to automation, trading greasemonkey scripts to fill forms, documenting ban rates by employer, and so on.
All this to say: I would be shocked if AI hadn’t had its hands in directly answering a large percentage of these answers.
Really interesting points.
Please do share the more interesting polls you have seen.
Full results here
72% saying AI made daily life better.
People only trust their doctor more to act in their best interest. The trust ChatGPT or similar over social media, elected officials, community leaders, faith leaders, and civil servants.
That seems incredibly foolish to trust the brand new shiny LLM over anything else. Not saying the rest can’t be untrustworthy but wtf?
Considering how they're algorithms designed to give satisfying-sounding answers at its core, even if you have stuff like guardrails and guidelines etc, I don't think it's that strange if you're not familiar with how they work.
In a sense producing a sense of trust is in their design, regardless or whether that's justified or not.
I totally expect that result. GPT and other LLMs are basically yes men that will speak in a reaffirming tone. Most people like that and build trust based upon it.
I remember feeling some kind of connection when I first used an LLM as well. Because at the time I didn't really understand how they worked, I was not mentally prepared for how much self-affirming language it would use. Nobody ever spoke to me like that, so it genuinely felt good (even if the output was telling me I was wrong, it told me in just about the nicest, most reassuring way I could imagine). I think that applies to a lot of people, we're so used to not having others reaffirm us that even a chat bot "love bombing" us ends up filling that void, for some more than others.
And that's the dangerous part, since chat bots don't know what they're saying, they're just regurgitating statistically probable answers, which puts many people in vulnerable positions now that all they listen to is their output.
I eventually I learned how they operate and now try to avoid them like the plague, but most people don't leave that first enamored phase, they simply keep using LLMs to reaffirm whatever they're already thinking and end up trusting the output a lot more as a result.
There's another strange thing happening with LLMs and how people don't have their guard up about them.
We are good at detecting things that are a little off and artificial. If these were robots with almost human faces, and they talked to you and looked at you but they just had soulless eyes or moved a little strangely or their skin looked a little plastic... And if they praised you too much and sometimes said really weird or wrong things... We wouldn't trust them at all and we'd be creeped out.
But it's just text so we are not properly suspicious. We haven't evolved enough yet to see threats that don't have a face I guess.
Well, let's break down the list, because it doesn't actually seem that foolish after really thinking about it.
Heavily dependant on the social media of course, but in general, I think I probably trust things I read less on social media than just random chance. It's virtually all ragebait and nonsense, so this one makes sense to me.
In the US, on a federal level, I think most of us can agree that most elected officials are literally professional con men at this moment in time. I tend to automatically assume whatever a federal official says is the exact opposite of the truth at this point and I'm usually proven correct. State and local vary by municipality, but in my case, it's mostly true for them as well.
This one is honestly surprising to me. I'm not religious or involved with community groups, but why would you be involved in those organizations if you didn't trust their leaders? They're voluntary for the most part
Highly dependent on most civil servants. For the most part, they're informed and professional, but I have had experiences where it was clear that they just wanted me to go away, and they were intentionally misleading me to get me lost in the gigantic cogs of bureaucracy. It doesn't happen to me often, because I'm basically what society assumes is a "default human being", a middle aged married straight cis white male with a steady job, no major health issues, and no criminal record. The further you deviate from that "default", the more problems I'd expect you'd have with civil servants.
Even though LLMs are going to hallucinate and affirm your stupid ideas, depending on circumstances, I'd trust its answer more than a lot of the people listed above. Doesn't seem that foolish to me overall.
I interpreted this from a slightly different angle. Think of "faith leader" not as your minister or pastor, but as "the ministers in my community, who are ostensibly trying to make thier communities better, in addition to preaching thier faith".
Even if the majority of people in a community "go to church", they don't all go to the same church, or even belong to the same faith. So it would be incorrect to assume that if someone goes to church and does trust their own minister, that doesn't necessarily mean they trust all of them.
The question was asking about trust in “your community or faith leader” so it is asking about the leader of your church specifically not just general community leaders
I (kind of) trust it more over all of those people, but only because I can repeatedly push back against it and direct it as more of a research partner than an authority, and I don't trust those people due to agendas, ignorance, laziness, or all three.
But I don't think most people actually push back and every time they show me their prompts, they aren't even asking the right question you would want to ask a real person, so it's a bit terrifying to me. But that's basically my takeaway anytime I encounter the thought processes of most people, so idk ¯\_(ツ)_/¯
I wonder what people are thinking of when they say AI. I think I'd agree with the statement that AI made daily life better, but I'd be talking about the ML algorithms that make modern search engines work well, the ability for my smart cameras to detect people instead of random shadows, my car's lane keeping assist functionality, a lot of the ML functions software I use at work uses to detect anomolies and so forth.
LLMs are largely negative, although I do find them useful sometimes. That doesn't really outweigh their content polluting search results though. They still have a pretty minor impact on my life other than supply chain shortages though.
If I take the sum of "AI" as a whole though, yeah, mostly positive.
I think one of the negatives of the term AI has always been its fuzzy definition which constantly changes in colloquial usage. I strongly suspect that people are only thinking about LLMs and diffusion based image generation models when they hear AI.
This is interesting for more than AI insights. It also looks into how people want the future shaped.
Ie
People prefer guaranteed work over guaranteed income.
It asks who people trust, with government right at the bottom.
I come from a country with guaranteed unemployment/disability income. Almost everyone much prefers guaranteed work. It pays so much better.
Ah, OK, I was curious about that! I only made it ~a third through the survey, but the point about people desiring to work vs. having freedom to do what they please, after their jobs have been rendered obsolete, kinda threw me. Maybe it's hard to imagine a world without the exploitation of labour, or as you (sideways) note, it's hard to imagine that there's enough for everyone such that we can all fully enjoy our lives without pretending that their 9-5 is critical to the functioning of society.
There is also a dignity factor. Many would prefer to work for their income rather than take handouts.
Personally I'd vote for income over work. It means that those who can't work are still taken care of. It also means that I'm not reliant on a workplace for survival. I'd hate to think what these guaranteed jobs are going to be like.
It sounds great to be able to do anything anytime. But it sounds pretty awful to have everyone else doing whatever they want at the same time as me! Plus, with no limitations, freedom can become its own prison. It becomes the hedonic treadmill.
Would you mind if I ask about why that is? I'm not sure if it's digging into anything personal -- so please feel free to ignore if that's the case -- but that sounds fantastic to me, imo. I love weird stuff like gigantic, steel spider sculptures appearing in underpasses (despite having a touch of arachnophobia), or music festivals popping up in random fields, so people being free to pursue their passions sounds lovely imo.
Q: aren't we basically talking about retirement, only early enough that peoples' bodies still function (and their friends are all still alive)?