i don't have a big audience here, but going to post this quick: working on a new word based 1v1 game: wordtrak.com - would love for some more beta testers to give me feedback!
05.03.2026 04:25
👍 4
🔁 2
💬 1
📌 0
i don't have a big audience here, but going to post this quick: working on a new word based 1v1 game: wordtrak.com - would love for some more beta testers to give me feedback!
nothing makes a microblogging site feel more like Twitter than arguing with libertarians
nah mine’s pretty great
if not I might try tomorrow
has anyone made a patreon labeler yet
don't worry it's thinking and actually will get the right answer eventually unless it doesn't then you're holding it wrong
419th goalpost mover brigade
yeah you should definitely trust this text generator to make sure your mental health is good
i already have an ai thread going lol but this is different - if you're a developer and not on linkedin, you are missing a lot of discourse from company execs and ai company dev evangelists talking about ai tokens being an employee perk up there with snacks and pto. this is v shady and v anti-worker
Cuba is in a major blackout now. No country brave enough to stop it. No people powerful enough to pressure their government. Bleak
I’ve been in therapy consistently for two decades :)
I wish. I just know that’ll never happen 😭
GitHub PR title: "V22.22.1 proposal #61947"
really hope we get one more patch release
literally just talked to the person sitting next to me about it and they said they tried it and both ChatGPT and Claude were too sycophantic to be a replacement for real therapy 😊
you know, it really makes sense that this kind of person sees nothing questionable about AI as a replacement for therapy 🙂↕️
you're the one putting your mental health in the hands of a text generator that makes shit up based off what reddit comments and 4chan think.
if my therapist was doing mushrooms and not connected to reality, I'd fucking leave their ass for malpractice in a heartbeat. hope you consider the same.
There are 3 "r"s in the phrase "hivemind revealer"
it literally says 3. It doesn't end at 3, but it does say 3. hiding this hallucination does not change that it hallucinated.
ICE’s violent lawlessness has no place in our communities.
Call your members of Congress now to urge them to block any bill that funds DHS without reining in their abuses.
your example also shows that it hallucinates - it *still* says 3. Mine says 4. Neither are correct.
I don't care that in a random example it can get to the right answer eventually, I care that it's objectively still hallucinating. Whether or not it can identify that in one example is irrelevant.
*hits bong* so it's really because they get to boss it around
I can also say things with no backing! Fictional characters are far more accurate and honest than AI lol
google search for "how many r's are in hivemindrivealer" says there are 4 r's
yeah? literally just typed this shit into google with zero effort and wow you're wrong
ah yes, LLMs are definitely known for always giving correct information
exactly this
rust programming language
how do you think mental illness gets treated?
and is it going to dissuade you or intervene when you're beginning to say some batshit crazy stuff? Or is it going to enable you because it's been prompted to be agreeable?
posted an hour ago btw bsky.app/profile/thev...
what do you mean they're not writing anything for you, you're literally advocating for them to replace a service that humans have developed.
"they're not writing anything for you"
"you have to read the output"
literally read your own words lmfao what are you reading
therapy should be free!
rather than giving your life information and money to the hypercapitalim machine, advocate for a better healthcare system that takes care of people in ways that *actually* help them.
those who committed suicide because they talked to LLMs deserved better.
