swlabr @ swlabr @awful.systems

Only Bayes Can Judge Me

---

Read more

Posts

38

Comments

1,815

Joined

2 yr. ago

Oh, that's a good angle too. Prompt the LLM with "what insights does this book have about B2B sales" or something.

take 2 minutes to think of precisely the information I need

I can’t even put into words the full nonsense of this statement. How do you think this would work? This is not how learning works. This is not how research works. This is not how anything works.

This part threw me as well. If you can think of it, why read for it? Didn’t make sense and so I stopped looking into this particular abyss until you pointed it out again.

I think the only interpretation of what this person said that approaches some level of rationality on their part is essentially a form of confirmation bias. They aren’t thinking of information that is in the text, they are thinking “I want this text to confirm X for me”, then they prompt and get what they want. LLMs are biased to be people-pleasers and will happily spin whatever hallucinated tokens the user throws at them. That’s my best guess.

That you didn’t think of the above just goes to show the failure of your unfeeble mind’s logic and reason to divine such a truth. Just kidding, sorta, in the sense that you can’t expect to understand an irrational thought process using rationality.

But if it’s not that I’m still thrown.

Dune

Prompts ChatGPT, skims the output because of muh ‘fishency

The Omelas Hole sure sounds like a paradise, I want to live in there!

That’s reasonable, and especially achievable if you don’t use chatbots or digital assistants!

I wonder if some of those chuds think those waymos might be conscious

what kind of semen retention scheme is this

Note to the Peanut gallery: this guy knows about paperclipmaxxing but not this more famous comic. Curious. lmfao

Incredible work as always, self

Dis you:

that being said… AI does literally help some people. for many things. google search was my favourite AI tool 25 years ago, but it’s definitely not right now.

lol

Lol, I’m a decision theorist because I had to decide whether I should take a shit or shave first today.

What's your P(doodoo)

time to donate my money to a different wiki that only has the noblest of intentions, wikifeet (jk)

Children really shouldn’t be left with the impression that chatbots are some type of alternative person instead of ass-kissing google replacements that occasionally get some code right

I agree! I'm more thinking of the case where a kid might overhear what they think is a phone call when it's actually someone being mean to Siri or whatever. I mean, there are more options than "be nice to digital entities" if we're trying to teach children to be good humans, don't get me wrong. I don't give a shit about the non-feelings of the LLMs.

hey dawg if you want to be anti-capitalist that’s great, but please interrogate yourself on who exactly is developing LLMs and who is running their PR campaigns before you start simping for AI and pretending like a hallucination engine is a helpful tool in general and specifically to help people understand complex topics where precision and nuance are needed and definitely not fucking hallucinations. Please be serious and for real

Very off topic: The only plausible reason I’ve heard to be “nice” to LLMs/virtual assistants etc. is if you are being observed by a child or someone else impressionable. This is to model good behaviour if/when they ask someone a question or for help. But also you shouldn’t be using those things anyhoo.

decidely unsexy. gerard is finally beating the allegations

Making LLMs safe for mentally ill people is very difficult

Arguably, they can never be made "safe" for anyone, in the sense that presenting hallucinations as truth should be considered unsafe.

For you, the day Disney graced your independent research lab was the most important day of your life. But for Disney, it was Tuesday.

-M. Bison (the M stands for Mickey Mouse)

If you want to hang out, you've gotta take her out, tech brain

If you want to get down, down on the ground, tech brain

AGI, say goodbye, ROI, tech brain

-Eric “Slophand” Slopton

Got curious and wanted to see if I could beat the Atari 2600. Found an online emulator here.

"Easiest" difficulty appears to be 8, followed by 1, then increasing in difficulty up to 7. I can beat 8, and the controls and visuals are too painful for me to try anything more than this.