4 ms·

And yet all of them use it...

Larrikin 5mo ago
Having Claude calculate which beers are the best deal at the bar based on price to alcohol from a picture of the menu is currently a massive party trick. Outside of programmers, almost no one has actually seen AI be useful for anything except do a barely acceptable job at a task they could have done better if they felt like it. Not all programmers with AI mandates have seen this yet either.
- simianwords 5mo ago
  It’s solving open problems in mathematics that stumped humans.
- mcv 5mo ago
  "If they felt like it" is key here: in my experience, AI makes it very easy to not feel like anything and just trust whatever the AI comes up with. The barrier to diving deep into code, or researching any other topic by myself, has become higher. At the same time the barrier to getting some results has become much lower, especially for complex topics I knew nothing about. So that seems great, but I keep running into cases where I do check sources and the AI turns out to have summarised it incorrectly.
  - arvid-lind 5mo ago
    It seems like that issue where LLMs are just flat wrong isn't going away, at least as long as they're trained on human data. Humans of course are fallible but it can be a cascading problem that isn't clear to anyone if all we're doing is looking at the solution and letting the LLMs check the work. I remember being excited about LLMs initially because it seemed like a natural evolution of the information saturated world we live in, we now need help deciphering the signals from the noise. But that was assuming that most/all the noise was from mis/disinformation. Now it just looks like another tool of control for the megarich to take even more from the rest of us. Maybe there is some hope in open source models.
tempaccount5050 5mo ago
No one uses it outside of tech or office jobs. There's no use case.
- jeroenhd 5mo ago
  I think it's worse than that. The biggest market for AI, possibly even bigger than tech, is mass manipulation, lying, and scamming. Destabilizing countries has never been easier now that social media and messengers allow believable lies and manipulation to spread like wildfire, and the AI industry has massively reduced the cost of believable lies. Up until a few years ago, believable videos of politicians or famous people or people targeted for blackmail were expensive and required acting or VFX work. Now anyone can do it with a handful of dollars and half an hour to spend. The industry is threatening to enrich the elite by taking people's jobs in economic uncertain times while at the same time resource hogging data centers are popping up all over the world like weeds. Big AI couldn't be more dislikable if they tried.
- simianwords 5mo ago
  False. I used it for 1. Filling bank forms, filling visa for South Africa 2. Understand movies and literature 3. For understanding public transport in new countries (pretty anxiety inducing) 4. As a 100x jump over Google search 5. Reading and answering emails 6. Fact checking dubious claims on the internet 7. Finding new music I might like Incurious read on AI belongs to say 2024.
  - ares623 5mo ago
    Imagine queueing up to get benefits and food stamps so simianwords can continue to find new music he might like.
    - simianwords 5mo ago
      Same could have been said during advent of internet. Imagine world has a fixed pie of wealth to distribute lol
    - 5mo ago
rsoto2 5mo ago
uh yeah, all the subscriber numbers from ai companies are because it was baked into every product humans already used on their tech i.e. browsers and search engines.
AngryData 5mo ago
Use it for... what exactly?
- s1artibartfast 5mo ago
  My favorite use is to give me PhD level tours of art museums and historic sites. It seems to know everything about every single artwork, the lives of the artists, and the economic and cultural context. It's willing to go at my pace and field as many or as few questions as I want. Frequently use it to come up with recipes when cooking, repair electrical equipment, or seek medical advice and results interpretation for my family. It's pretty hard to imagine life without it at this point. I know it's possible, but like the internet, I would feel crippled by the lack of information and things that I can no longer easily do
  - scubadude 5mo ago
    What about when it just flat out lies to you, how do you sort the right info from the wrong? > recipes when cooking I used it for a recipe, gave it a brilliant and detailed prompt, it told me to put 10x a particular spice and it ruined the dish. > seek medical advice and results interpretation for my family good luck > repair electrical equipment what can go wrong, really
    - simianwords 5mo ago
      Why do you use hacker news when anyone can lie to you? You still use it because it is directionally correct and the accuracy rate is high enough that it makes it worth it. Same with ChatGPT. I use the thinking model and rarely (if not never) get obvious errors.
    - s1artibartfast 5mo ago
      Its not perfect, just like talking to humans, online information, the news, or even books. It makes mistakes all the time, so you have to think and judge, just like the rest of life, but I find it is a tremendous resource. Recipies are one of the strongest areas given the low complexity and obvious training data. Im curious what it messed up on with.
LastTrain 5mo ago
Whether they want to or not even

Having Claude calculate which beers are the best deal at the bar based on price to alcohol from a picture of the menu is currently a massive party trick. Outside of programmers, almost no one has actually seen AI be useful for anything except do a barely acceptable job at a task they could have done better if they felt like it. Not all programmers with AI mandates have seen this yet either.

simianwords 5mo ago
It’s solving open problems in mathematics that stumped humans.
mcv 5mo ago
"If they felt like it" is key here: in my experience, AI makes it very easy to not feel like anything and just trust whatever the AI comes up with. The barrier to diving deep into code, or researching any other topic by myself, has become higher. At the same time the barrier to getting some results has become much lower, especially for complex topics I knew nothing about. So that seems great, but I keep running into cases where I do check sources and the AI turns out to have summarised it incorrectly.
- arvid-lind 5mo ago
  It seems like that issue where LLMs are just flat wrong isn't going away, at least as long as they're trained on human data. Humans of course are fallible but it can be a cascading problem that isn't clear to anyone if all we're doing is looking at the solution and letting the LLMs check the work. I remember being excited about LLMs initially because it seemed like a natural evolution of the information saturated world we live in, we now need help deciphering the signals from the noise. But that was assuming that most/all the noise was from mis/disinformation. Now it just looks like another tool of control for the megarich to take even more from the rest of us. Maybe there is some hope in open source models.

tempaccount5050 5mo ago

No one uses it outside of tech or office jobs. There's no use case.

jeroenhd 5mo ago
I think it's worse than that. The biggest market for AI, possibly even bigger than tech, is mass manipulation, lying, and scamming. Destabilizing countries has never been easier now that social media and messengers allow believable lies and manipulation to spread like wildfire, and the AI industry has massively reduced the cost of believable lies. Up until a few years ago, believable videos of politicians or famous people or people targeted for blackmail were expensive and required acting or VFX work. Now anyone can do it with a handful of dollars and half an hour to spend. The industry is threatening to enrich the elite by taking people's jobs in economic uncertain times while at the same time resource hogging data centers are popping up all over the world like weeds. Big AI couldn't be more dislikable if they tried.
simianwords 5mo ago
False. I used it for 1. Filling bank forms, filling visa for South Africa 2. Understand movies and literature 3. For understanding public transport in new countries (pretty anxiety inducing) 4. As a 100x jump over Google search 5. Reading and answering emails 6. Fact checking dubious claims on the internet 7. Finding new music I might like Incurious read on AI belongs to say 2024.
- ares623 5mo ago
  Imagine queueing up to get benefits and food stamps so simianwords can continue to find new music he might like.
  - simianwords 5mo ago
    Same could have been said during advent of internet. Imagine world has a fixed pie of wealth to distribute lol
  - 5mo ago

rsoto2 5mo ago

uh yeah, all the subscriber numbers from ai companies are because it was baked into every product humans already used on their tech i.e. browsers and search engines.

AngryData 5mo ago

Use it for... what exactly?

s1artibartfast 5mo ago
My favorite use is to give me PhD level tours of art museums and historic sites. It seems to know everything about every single artwork, the lives of the artists, and the economic and cultural context. It's willing to go at my pace and field as many or as few questions as I want. Frequently use it to come up with recipes when cooking, repair electrical equipment, or seek medical advice and results interpretation for my family. It's pretty hard to imagine life without it at this point. I know it's possible, but like the internet, I would feel crippled by the lack of information and things that I can no longer easily do
- scubadude 5mo ago
  What about when it just flat out lies to you, how do you sort the right info from the wrong? > recipes when cooking I used it for a recipe, gave it a brilliant and detailed prompt, it told me to put 10x a particular spice and it ruined the dish. > seek medical advice and results interpretation for my family good luck > repair electrical equipment what can go wrong, really
  - simianwords 5mo ago
    Why do you use hacker news when anyone can lie to you? You still use it because it is directionally correct and the accuracy rate is high enough that it makes it worth it. Same with ChatGPT. I use the thinking model and rarely (if not never) get obvious errors.
  - s1artibartfast 5mo ago
    Its not perfect, just like talking to humans, online information, the news, or even books. It makes mistakes all the time, so you have to think and judge, just like the rest of life, but I find it is a tremendous resource. Recipies are one of the strongest areas given the low complexity and obvious training data. Im curious what it messed up on with.

LastTrain 5mo ago

Whether they want to or not even