Plik's recent activity
-
Comment on Crypto's new bailout fund: your savings account in ~society
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik Link ParentYes. But I am mostly seeing comments about how people expect AI companies to do the right thing. ...Why? History kinda proves a lot of companies don't do that unless there's a profit incentive, or...Either way, the expectation, IMO is for governments to actively hold the corporations accountable.
Yes. But I am mostly seeing comments about how people expect AI companies to do the right thing.
...Why? History kinda proves a lot of companies don't do that unless there's a profit incentive, or government oversight.
And it's not letting them off the hook, it's just...stop blaming a thing for behaving the way that thing was designed to act, and look for actual solutions instead of whining on the internet. If you actually care enough to do anything about it. Otherwise it's just concern performance art.
Edit: To answer your question about corporate drive, corporations are driven by profit, that's it, anything else is window dressing to make people feel good about...the company's profit. The only other option is government. But there's probably no hope there for the next ~3.5 years.
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik Link ParentI mean, I would blame TikTok influencers and society/government first. But 🤷🏻♂️. Easier to blame faceless corporate America. People and thus corporations can't always be trusted to do the morally...I mean, I would blame TikTok influencers and society/government first. But 🤷🏻♂️. Easier to blame faceless corporate America.
People and thus corporations can't always be trusted to do the morally correct thing, and I think trusting them to do so is foolish. It's gotta come from everyone putting pressure on their governments to enforce doing the right thing for society as a whole.
And yes, I know that doesn't always work the best.
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik Link ParentNo. Both can be true. Maybe ChatGPT is sending me Onion article derived info and I don't know. That doesn't mean it hasn't had an effect on my thought processes. And yes it always in the back of...No. Both can be true.
Maybe ChatGPT is sending me Onion article derived info and I don't know. That doesn't mean it hasn't had an effect on my thought processes. And yes it always in the back of my mind that it may/is feeding me bs, which I why I believe only adversarial interactions allow it to be useful, along with confirming whatever it says with outside sources.
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik Link ParentI get that, but also you can apply it to the most innocent of things too. Too much protein powder = kidney problems. You blame the company or the consumer or society? Too much makeup/improper...I get that, but also you can apply it to the most innocent of things too. Too much protein powder = kidney problems. You blame the company or the consumer or society?
Too much makeup/improper removal = pimples and skin problems. Is that the company or the user's fault?
Car can do 200 kph? Driver skids off the road?
A lot are below average, but at some point you can't keep blaming the companies. I am in general a fan of AI now after learning how to use it, not a fan of the companies though. But at the same time where is the line where it is no longer their fault if the AI is misused?
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik (edited )Link ParentUgh, I'll bite. What? No, I never said that, and it's a huge logical jump to somehow insinuate that using an LLM is slavery, or akin to, or on the path to? Also, I don't think you can just claim...Ugh, I'll bite.
When you say AGI, because these words fundamentally have no meaning anymore, do you mean a person? Because at some point we're doing slavery while pretending it's different. Asimov's robots are slaves.
What? No, I never said that, and it's a huge logical jump to somehow insinuate that using an LLM is slavery, or akin to, or on the path to? Also, I don't think you can just claim the words fundamentally have no meaning anymore, who decided that? You? . I love sci-fi, but that is an absurd take.
I mean AGI in the sense that it can solve problems generally without human guidance that only well qualified humans have been capable of until now-ish. AI as we currently have it needs human guidance and feedback...I was never talking about sci-fi space AI, and it's weird to try and twist my words into saying that I was talking about AI slavery.
Is it the lay persons fault for misunderstanding when the corporation intended you to?
After some point maybe? I don't believe it's a hard no. If corporations are bad, shouldn't even "lay people" learn to distrust them? This isn't a "the AI company tricked you into a time share once" kind of thing, it would be repeatedly going against your instincts, and ignoring your family/friend's advice in favor of the AI's. At some point people need to take at least some responsibility for their actions.
You're using AI, why do you feel the need to use it to the extent you don't know if you've been fundamentally changed by it or not?
I don't feel the need, I have just used it so much for a few specific tasks that it has changed me (significantly), and it feels very strange and different than how books, teachers, textbooks, friends, and family influence one's life. The machine is not alive, but I am (I think?), but the machine has changed my thought processes, so am I now part machine?
That's a fairly extreme statement. Why don't you seek human support and advice?
Because the topics I am asking about are mostly reserved to the realm of applied math/computational finance/physics PhDs...or funds that have billion dollar buy ins. There is nobody near me to talk about these things with, and the internet is full of fake gurus, gamblers, and scammers.
Why do you feel the need to go to the bot and ask it to be mean to you, explicitly?
Because I need a rubber duck, not a sycophant. Part of this thread has already brought up concerns about AI being too nice and where that leads. Easy solution? Tell it to be mean (it will still be too nice, just less so, until you ask ~3 times for more direct and mean counter points). This isn't some bizarre be mean to me because I hate myself and want to suffer, it's literally just the machine being far too nice, and having to use a two meter long lever to get it turned down from 11.
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik Link ParentYeah, my "that's on you" statement was directed at general use while ignorantly ignoring the context. I guess what I was trying to get at and failed was that I think AI is incredibly useful if...Yeah, my "that's on you" statement was directed at general use while ignorantly ignoring the context.
I guess what I was trying to get at and failed was that I think AI is incredibly useful if used correctly (which would be hard to do in the wrong mental state), but also that I don't think blaming AI and the companies gets at the root of the problem. To me it's similar to blaming a (legitimate) supplement company for someone taking too many supplements (e.g. body dysmorphia and consuming too much protein powder resulting in kidney issues). I just don't think it's competely reasonable to blame the product over other factors.
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik Link ParentYeah, that makes sense. Weird to me personally because I usually assume everything might be an Onion article.Yeah, that makes sense. Weird to me personally because I usually assume everything might be an Onion article.
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik (edited )Link ParentFair enough. Edit: So while it was tone deaf, I still feel that blaming the AI/company is the easy way out. I think the real question is why does the person feel the need to go to the AI?...Fair enough.
Edit: So while it was tone deaf, I still feel that blaming the AI/company is the easy way out. I think the real question is why does the person feel the need to go to the AI? Shouldn't there be some sort of societal support before it even gets there?
My new changed opinion on AI/AGI is that it could actually be a sort of "ultimate" democratizer. The tech bro corporate dystopia is more of a directed future (by the tech bros). What if we were somehow able to create personal (non-corporate owned) open source AGI that everyone could use? Sort of like Asimov's androids. Everyone with their own AGI to help protect them from the greed and selfishness in the world?
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik Link ParentSlightly, but not much different than an Andrew Tate sub-reddit with terminally online incels doing feedback loops on each other.Slightly, but not much different than an Andrew Tate sub-reddit with terminally online incels doing feedback loops on each other.
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik Link ParentTrue, but that also kind of applies to the entire internet 😅True, but that also kind of applies to the entire internet 😅
-
Comment on Everyone is cheating their way through college in ~tech
-
Comment on Everyone is cheating their way through college in ~tech
Plik Link ParentThis is the only way. Most of the grade needs to be determined by work done in front of a trusted observer. And as bad as it may sound, students need to be put on the spot more so they can...This is the only way. Most of the grade needs to be determined by work done in front of a trusted observer.
And as bad as it may sound, students need to be put on the spot more so they can experience how cheating with AI will screw them in the moment.
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik Link ParentI literally ask it "is that the mean version?" whenever I think it's blowing smoke up my ass, and if it says no, I tell it to give me the mean version. That usually works, and has actually crushed...I literally ask it "is that the mean version?" whenever I think it's blowing smoke up my ass, and if it says no, I tell it to give me the mean version. That usually works, and has actually crushed my ego a few times (a good thing IMO).
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik LinkSo here's the thing...you can ask it to disagree with you, or take the other side and it will. This is one of its best uses IMO. I have basically been converted to the side of the ...Thronglings....So here's the thing...you can ask it to disagree with you, or take the other side and it will. This is one of its best uses IMO.
I have basically been converted to the side of the ...Thronglings.
To me, if you aren't asking it to be adversarial, you're using it wrong, and that's on you. Just the same as someone who only sticks to niche echo chamber sub-reddits rather than looking for challenges to their beliefs, needs to own the fact that they have chosen the comfortable discourse that makes them feel better, rather than the discourse that challenges their intellect.
-
Comment on Some ChatGPT users are developing delusional beliefs that are reinforced by the large language model in ~tech
Plik Link ParentI use ChatGPT extensively, and it's not exactly hype. You have to be very careful with it, but with careful use it can be extremely helpful. I actually had a bit of an existential...I use ChatGPT extensively, and it's not exactly hype. You have to be very careful with it, but with careful use it can be extremely helpful.
I actually had a bit of an existential realization/crisis last night in that I realized ChatGPT has completely changed certain aspects of...me (for the better, I think?).
And now I am deeply affected about whether I am "me" anymore, or some weird human/ChatGPT hybrid 😅
-
Comment on Business idea and feedback thread in ~talk
Plik (edited )Link ParentOnline is probably not the best way to go IMO, you'd be competing with absolutely massive online English companies. If you want to do corporate English, then that may work, but would probably be...Online is probably not the best way to go IMO, you'd be competing with absolutely massive online English companies.
If you want to do corporate English, then that may work, but would probably be better in person, especially helping with presentations, meetings, and writing emails....Although AI can pretty much do the third one for anyone.
Each country has their own giant English center monstrosity that kinda takes over the in person stuff unfortunately, so this would take a lot of work to make it unique.
Good thing is, once you get a few fancy clients who like you, they will send you more work for organization training events, which can pay very well.
Large corporations, governments, and NGOs tend to overpay if you do group training.
I did this for an agency when I was younger, and another friend hit the jackpot with a corporate gig. Thing is jackpot generally means good hourly rate, but not many + infrequent hours.
-
Comment on Business idea and feedback thread in ~talk
Plik Link ParentI know nothing about restaurants, but I think keeping the menu super compact is a good idea. In Asia you often have street stalls where they literally make one thing. Then they maybe have cheap...I know nothing about restaurants, but I think keeping the menu super compact is a good idea. In Asia you often have street stalls where they literally make one thing. Then they maybe have cheap condiments you can mix in to adjust the flavor to taste (self-serve).
-
Comment on End of 10: Replace Windows 10 with Linux in ~tech
Plik Link ParentIt was the first time I used Windows and actually liked it. I prefer Linux....but also want normal PC things to just work with almost zero hassle.It was the first time I used Windows and actually liked it. I prefer Linux....but also want normal PC things to just work with almost zero hassle.
-
Comment on End of 10: Replace Windows 10 with Linux in ~tech
Plik Link ParentIt gets expensive for a "real" Windows machine. I am running the lowest tier I can on AWS to run one program, and I start and stop the instance when not using it, t3.xlarge is about $0.35/hour and...It gets expensive for a "real" Windows machine. I am running the lowest tier I can on AWS to run one program, and I start and stop the instance when not using it, t3.xlarge is about $0.35/hour and maybe $2/month for 35 GB of storage (charged whether or not the instance is running). I tried t3.large but it wasn't enough CPU to run anything useful, 100% CPU just for the OS and one program.
I don't know how much Windows Server costs for...a real computer.
However WS2025 is like...impressively not bloated, reminds me of the old MacOS where you got an OS and nothing more.
99.9% Blackrock is taking over it all. ETHA, IBIT, CRCL (USDC), Blackrock is behind all of them.
XLF has out performed SPY by almost 2x (13% vs 24%) in the last year.
I think it is a planned move to US pegged USDC, and building a smart contract based system (ETHA), to basically replace SWIFT, IBAN, Mastercard, and Visa.
Cut out the banks, create a global payment infrastructure, and track everything. Throw in AI and you have a rather dystopian financial future.
Also Robinhood just bought Bitstamp.