561–570 of 1001 posts
Earlier quoted context omitted.
Serious question - why Texas???
It's where the energy is for this project. This is unfortunately paywalled but a good writeup on how the datacenter came to be: https://www.theinformation.com/articles/why-openai-and-oracl...
I'm confused and a bit disturbed; honestly having a very difficult time internalizing and processing this information. This announcement is making me wonder if I'm poorly calibrated on the current progress of AI development and the potential path forward. Is the key idea here that current AI development has figured out enough to brute force a path towards AGI? Or I guess the alternative is that they expect to figure…
> Is the key idea here that current AI development has figured out enough to brute force a path towards AGI? My sense anecdotally from within the space is yes people are feeling like we most likely have a "straight shot" to AGI now. Progress has been insane over the last few years but there's been this lurking worry around signs that the pre-training scaling paradigm has diminishing returns. What recent outputs like…
My problem with this is that people making this statement are unlikely to be objective. Major players are in fundraising mode, and safety folks are also incentivised to be subjective in their evaluation.
Yesterday I repeatedly used OpenAI’s API to summarise a document. The first result looked impressive. However, comparing repeated results revealed that it was missing major points each time, in a way a human would certainly not. In the surface the summary looked good, but careful evaluation indicated a lack of understanding or reasoning.
Don’t get me wrong, I think AI is already transformative, but I am not sure we are close to AGI. I hear a lot about it, but it doesn’t reflect my experience in a company using and building AI.
https://situational-awareness.ai/racing-to-the-trillion-doll...
Earlier quoted context omitted.
"There are maybe a few hundred people in the world who viscerally understand what's coming. Most are at DeepMind / OpenAI / Anthropic / X but some are on the outside. You have to be able to forecast the aggregate effect of rapid algorithmic improvement, aggressive investment in building RL environments for iterative self-improvement, and many tens of billions already committed to building data centers. Either we're a…
Motivated reasoning sings nicely to the tune of billions of dollars. None of these folks will ever say, "don't waste money on this dead end". However, it's clear that there is still a lot of productive value to extract from transformers and certainly there will be other useful things that appear along the way. It's not the worst investment I can imagine, even if it never leads to "AGI"
A 500B dollar investment doesn't just fall into one's lap. It's not your run of the mill funding round. No, this is something you very actively work towards that your funders must be really damn convinced is worth the gamble. No one sane is going to look at what they genuinely believe to be a dead end and try to garner up Manhattan Project scales of investment. Careers have been nuked for far less.
Earlier quoted context omitted.
"There are maybe a few hundred people in the world who viscerally understand what's coming. Most are at DeepMind / OpenAI / Anthropic / X but some are on the outside. You have to be able to forecast the aggregate effect of rapid algorithmic improvement, aggressive investment in building RL environments for iterative self-improvement, and many tens of billions already committed to building data centers. Either we're a…
My prediction is a Apple loses to Open AI who releases a H.E.R. (like the movie) like phone. She is seen on your lock screen a la a Facetime call UI/UX and she can be skinned to look like whoever; i.e. a deceased loved one. She interfaces with AI Agents of companies, organizations, friends, family, etc to get things done for you (or to learn from..what's my friends bday his agent tells yours) automagically and she is…
Earlier quoted context omitted.
Data centers are nearly all blue collar work.
If you're familiar with this kind of work, please elaborate! Do you mean building the centers or maintenance or both?
Once you have one working design for the environment (e.g. hot desert vs cold and humid), you can stamp the things out with minimal variation between the two.
The maintenance of all of that supporting infrastructure is the standard blue collar work the same.
The only new blue collar job on the maintenance side is responding to hardware issues. What this entails depends on if it’s a colo center and you’re doing “remote hands” for a customer where you’re swapping a PSU, RAM, or whatever. You also install new servers, switches, etc.
As you move up into hyperscalers the logistics vary because some designs make servicing a single server in place not worth cooling the whole hot aisle (Google runs really hot hot aisles that weren’t human friendly). So sometimes you just yank the server and throw it in a cart or wait for the whole rack to fail and pull it then.
Overall though, anything that can be done remotely is. So the data center techs do very little work on the keyboard
Oh so that's why Pelosi invested in Micro nuke electricity plants.
The biggest question on such investment from my POV, is what do the Deepseek results mean about the usefulness/efficiency of this investment? I've been meaning to read a relevant book to today's times called Engines That Move Markets. Will probably get it from the library.
For fun, I calculated how this stacks up against other humanity-scale mega projects. Mega Project Rankings (USD Inflation Adjusted) The New Deal: $1T, Interstate Highway System: $618B, OpenAI Stargate: $500B, The Apollo Project: $278B, International Space Station: $180B, South-North Water Transfer: $106B, The Channel Tunnel: $31B, Manhattan Project: $30B Insane Stuff.
It's unfair, because we are talking in the hindsight about everything but Project Stargate, and it's also just your list (and I don't know what others could add to it) but it got me thinking. Manhattan Project goal is to make a powerful bomb. Apollo is to get to the Moon before soviets do (so, because of hubris, but still there is a concrete goal). South-North Water Transfer is pretty much terraforming, and others ar…
It has been quite clear for a while we'll shoot past human-level intelligence since we learned how to do test-time compute effectively with RL on LMMs (Large Multimodal Models).