ChatGPT has a word length limit of around 1000 words per time which is pretty pathetic. It also will after awhile start a new chat with no memory of the old one. It’s AI art is limited to around 8 per day and it puts each idea out in pairs. Yes, I was a paid member i Just cancelled it. Do things like ask it for a representation of a person writing with a left hand and it keeps coming back with right hands.
Probably because it’s coming back with the mean answer and not outliers.
And each bad reponse is charged against the limit. I recently dropped my paid account. My employer has an inhouse chatbot that is running on an older version of chatGPT, but it won’t ouput files.
Or ask it to show a train coming off the rails and it keeps the wheels firmly planted for 3 iterations until you yell at it:-). Sometimes it tells me it can’t do something but then it can when I basically remind it that it can. And that gets charged as 2 prompts when only one response was useful. Or however they count prompts/reponses.
https://www.nature.com/articles/s41467-024-54178-1
Having AI design something that works and have no idea why it works. The holy grail of engineering.
This is definitely an issue on YouTube and Facebook and also some other sites I use.
Does anyone know the best AI for marketing, Such as finding out which countries to Target for certain interests.?
Another interesting article suggesting capital will matter more than ever in a post AGI world…
Labour-replacing AI will shift the relative importance of human v non-human factors of production, which reduces the incentives for society to care about humans while making existing powers more effective and entrenched
Just received an e-mail from a guy who would like to cooperate with us.
The letter was most certainly written by AI and it’s such a mouthful of hyperbole. Hard to digest and totally inappropriate, cause it misses any personal touch that might make me interested in responding to this person.
Just another example of AI going down the wrong path.
ChatGPT has been awful lately
Me: do this relatively simple thing
Chatgpt: sure, I can fuck that up for you
Me: no, you didn’t do it correctly. Like i told you, try again
Chatgpt: oh, you’ve used up your free use for today, try again tomorrow!
That’s how they move you to be a paying customer. How about the other platforms ?
I guess i should try one 
Which do you recommend?
I dunno, I saw there’s a Chinese one that’s just as good, deepseeker. Gemini. What else ? Grok ?
That is actually how they lost me as a paying customer.
Can I ask how your experience has been with Aligner? I “passed” an AI interview, which was completely… weird, I guess, and am hesitating to go through with the on-boarding, which requires a background check. I saw some negative comments on reddit but can’t really tell if the complaints are legitimate.
“Largely positive”, I would say. I got invited there from Upwork back in October. They’re an enterprise client on Upwork, and I think they basically invited a ton of people with PhDs in STEM topics, as well as people working in other subjects like languages.
That said, so far I’ve only done like 6–7 hours of work on a single short project with them, and that project is currently paused. It was well paid — I’ve made US$950 or thereabouts up to now for that small amount of work — so I’m not complaining too much, but from the Slack and Discourse channels it looks like many people haven’t gotten any work at all.
The strategy with these companies seems to be to onboard as many people as possible for projects that might come up. So when you get work you get work and it’s fine, but I wouldn’t treat it as a regular source of income at all. I also wouldn’t expect them to be great at responding until there’s a project for you. Quite frustrating, but that’s how it is.
I haven’t done the AI interview yet. That’s quite recent, and there was some technical issue for ages where it wasn’t showing up for people coming from Upwork. It just appeared for me a couple of days ago and I haven’t gotten around to it yet. I’ll probably do it soon.
Anyway, I’d definitely recommend doing the onboarding and seeing how it pans out. It’s really not that much hassle — probably just identity verification, watching a couple of videos, and reading some stuff.
I’d recommend Outlier too. No idea how long it’ll last, but I’m getting quite a consistent stream of work from them at the moment and the project I’m on is fairly interesting (annotating and improving chemistry/biology/physics prompts involving complex reasoning stuff).
Best advice if you do get a project (on either platform) is probably just to read the instructions carefully and do your best to follow the rules and do a good job. I got promoted to a reviewer on the above project pretty quickly after just a handful of tasks, so I’m seeing all the low-quality junk many people submit. These platforms have to deal with quite a lot of spammers, AI-generated submissions (universally against the rules), people selling accounts so unqualified people can submit tasks, and just your common or garden people doing shitty work, so I think it’s pretty easy to get a decent source of income if you’re reasonably competent and careful.
Thanks for the thoughtful reply, Andrew! I applied directly with Alignerr. I don’t have an advaced degree but do have a couple decades of experience in the STEM area they were looking for. The AI interview was weird, not only because I was talking to a faceless AI bot while staring at my own face in the monitor–very disturbing–but also because the focus of the questions was my background in management and strategy, not the subject area I had expressed interest in. Nothing technical at all. That interviewer AI would be a great choice for voice actor if they ever remake “2001: A Space Oddyssey” with a female HAL 9000.
Sounds like its not a scam, so will give it a try. I am interested in AI, so this could be a good learning experience if I actually get a project. A few hours every couple weeks would be ideal. I will take a look at Outlier, too.
No worries, and good luck!
Where are you by the way? Last time I checked a month or so ago with a friend, I wasn’t sure it was possible to apply directly from Taiwan (it wasn’t listed as a country option for the position we were looking at, at least).
I don’t think an advanced degree is necessarily required for some of this stuff, if the person has experience like you said you have. I suspect they just used the PhD thing as a basic filter for people who might be qualified.
I’ve heard that from someone else about the AI interview! From what I gather, the system is supposed to automatically match people to suitable projects based on their experience and resume, so I guess the weird interview is probably just one part of that.
I think it’s definitely not a scam, just a bit chaotically organized and inconsistent work. It’s quite fun to work on too! And the models are getting better and better. It’s getting really rather challenging to write prompts the models can’t solve (which is the point of the Outlier project I’m working on).
Could you sign up a bunch of AIs with ‘PhDs’ to do this work on your behalf ?
All of the platforms I’ve signed up for or worked on have pretty strict rules on duplicate/fake accounts, spam submissions, use of AI, plagiarism, etc.
I’m sure it still happens, but they do try quite hard to stop it. I’ve reported a fair few submissions myself because I suspected AI use or otherwise gaming the system.
The point is they’re wanting reinforcement learning from human feedback (RLHF), so obviously they specifically want humans doing the work rather than other AI models (at least until we become obsolete).
I am in California. Hoping to semi-retire from the main career in a year or two and possibly move back to Taiwan.