You might like
I know this is sarcasm but this generalized attitude is def a prob - what happens once all that matters is your bag? Expertise is overrated/obsolete they say?
You: knowing how operating systems work while you change the color of a button for your PM Me: pushing my uber for dogs code that is incomprehensible to me which I had GPT4 write and then selling the app to byte dance for a bag
The counter factual is there is no “doom” or foom and we have deliberately crippled our own capacity because some people liked the terminator movie too much. Or even darker we have given all the reigns to mega corps. Individuals lose power.
You are right that open source brings in different challenges in relation to AGI extinction risk, and there's no easy solutions that will make everyone happy (as is usual in life, alas). The OSS community (which I love deeply) has a double-edged sword of a culture of…
Imagine being so terminally online that you believe the entirety of human experience has been encoded into text and published on the internet. #chatgptlol
Oh goodie the ai companies are trying to raise the regulatory moat. Sounds like everyone gets crappy ai except for certain well connected people
"If AI systems wanted to push humans out, they would have lots of levers to pull" "The trend will probably be towards these models taking on increasingly open-ended tasks on behalf of humans. The culmination of this is the “obsolescence regime”: for any task, you would rather…
It's telling to me that doomers can't, say, point at a reasonable 10 or even 50 page summary of their arguments that is sufficiently thorough, self-contained, and falsifiable that one can know whether one has or has not successfully rebutted all their arguments.
Okay I buy this explanation. A mistaken belief that cannot be corrected by his staff, well that’s been on full display
"If AI systems wanted to push humans out, they would have lots of levers to pull" "The trend will probably be towards these models taking on increasingly open-ended tasks on behalf of humans. The culmination of this is the “obsolescence regime”: for any task, you would rather…
The only problem with the Turing test is how gullible people are, especially smart people. Like some chat bot says “don’t kill me” and bam you think the matrix math is alive? Like only if your model of cognition is very abstracted or simplistic!
This is the kind of overly credulous behavior in “smart” people that is rightly decried. Come on man, companies don’t have secrets and rarely see the future better than anyone else.
It is hard to figure out whether: A) OpenAI can see the future more clearly then us, because they are working on future AIs B) OpenAI is overly optimistic, because many startups building tech are too optimistic C) This is marketing But this line in their blog is worth pondering.
There was indeed Walmart in 1998 and they were in full swing crushing small town America
It is hard to figure out whether: A) OpenAI can see the future more clearly then us, because they are working on future AIs B) OpenAI is overly optimistic, because many startups building tech are too optimistic C) This is marketing But this line in their blog is worth pondering.
I just taught my 5yo that a “4 double bunk bed” is a quadruple bunk bed and he said “that’s the baddest name ever”, and I can’t say he’s wrong!
Sounds like they had a hard time with all those colonies, really in many ways they were doing Africa a favor, right?
This is your periodic reminder that colonization was a money pit for France and is obviously not what made it rich, but rather it's because it was rich that it was able to colonize Africa. Today French trade with Africa is tiny and we're a net exporter of food products to Africa.
So I just walked about 5-6 blocks in western soma and got on the Bart train at civic center. I have never in my life seen less homeless people. I saw 0 panhandlers and maybe 2 homeless people on my 20 minute walk. Bart train is new, clean, and again no homeless. #sf
We’ve had this argument about “access to information is dangerous” for decades. Turns out actually doing things is really hard. Unlike, for example, writing this paper.
"The ‘Safeguarding the Future’ course at MIT tasked non-scientist students with investigating whether LLM chatbots could be prompted to assist non-experts in causing a pandemic. In 1 hour, the chatbots: - suggested 4 potential pandemic pathogens - explained how they can be…
So, are they sure such a theory or the start of one doesn’t exist? There’s a lot of work in psychology, education, and philosophy that might have something to say about this. Or maybe just read some great literature?
a lot of the future may hinge on whether we have a good theory of consciousness by the time agi gets here
GIGO - garbage in, garbage out. If you put infinitesimals in a 1/x then highly remote things become practically guaranteed to happen. It cloaks itself in the pretend rigor of math to cosplay as science.
i 100% do not understand how people are calculating probabilities of x-risk. it all seems like made up numbers on top of weird unchecked assumptions on top of too many SF dinner parties (not unlike ML). i've made a half-assed attempt to try to understand but that made it worse
It’s also known as anchoring - by calling out a specific value it sets the point at which people think. That way when the line is redrawn to $200k it seems totally reasonable.
The point of the "I make $400k and can't pay my bills" articles is purely to generate clicks off of rage-baiting normal people who make much, much less than $400k and still pay their bills while getting the article shared by starting discourse about where the line for "rich" is
Ah yes science fiction I love it
If transformers actually can reach AGI we should expect an intervention in the next 24-36 months imo. Some possibilities: - sophon cap - knowledge/tech transfer to solve alignment - destruction w/o warning (unlikely imo) - destruction w/ warning (motive for UAP cover-up)
United States Trends
- 1. Good Thursday 24.9K posts
- 2. Happy Friday Eve N/A
- 3. #thursdayvibes 1,898 posts
- 4. #Talus_Labs N/A
- 5. #thursdaymotivation 1,150 posts
- 6. Luka 72.8K posts
- 7. #ThursdayThoughts 1,203 posts
- 8. Lakers 83.2K posts
- 9. #River 4,384 posts
- 10. Wemby 28.1K posts
- 11. Vatican 10.8K posts
- 12. Unplanned 6,113 posts
- 13. Marcus Smart 6,682 posts
- 14. Captain Kangaroo 1,224 posts
- 15. Blazers 9,582 posts
- 16. Russ 11.7K posts
- 17. Richard 46.3K posts
- 18. Ayton 17.8K posts
- 19. Shabbat 7,136 posts
- 20. Shroud 5,835 posts
Something went wrong.
Something went wrong.