Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AGI fantasy is a blocker to actual engineering

tomwphillips.co.uk

Post reply on HN
Print view View original

371–380 of 683 posts

Re: AGI fantasy is a blocker to actual engineering

#371

Post by sumeno »

simianwords wrote: post #294
danaris wrote: post #137

Earlier quoted context omitted.

Yes, it is worthwhile to ask how much value we get. And a whole bunch of us are saying we don't see the value in all these datacenters being built and run at full power to do training and inference 24/7, but you just keep ignoring or dismissing that. It is absolutely possible that generative AI provides some value. That is not the same thing as saying that it provides enough value to justify all of the resources bein…

it is also a fraction of golf courses which you again ignore. this is just typical "don't do anything!!" ism. there's no argument here.. even if data centres used .00001 millilitre of water you would say the same thing.

Complaining about someone else using an "ism" during a literal "whataboutism" is rich

Re: AGI fantasy is a blocker to actual engineering

#372

Post by danaris »

simianwords wrote: post #294
danaris wrote: post #137

Earlier quoted context omitted.

Yes, it is worthwhile to ask how much value we get. And a whole bunch of us are saying we don't see the value in all these datacenters being built and run at full power to do training and inference 24/7, but you just keep ignoring or dismissing that. It is absolutely possible that generative AI provides some value. That is not the same thing as saying that it provides enough value to justify all of the resources bein…

it is also a fraction of golf courses which you again ignore. this is just typical "don't do anything!!" ism. there's no argument here.. even if data centres used .00001 millilitre of water you would say the same thing.

Oh, I think golf courses shouldn't exist. They're awful in a number of ways. You want to play golf? VR or minigolf.

But (as I pointed out elsewhere in this discussion [0]) why should I have to mention everything that uses water in a way I think is detrimental in order to be allowed to talk about why I think this thing uses water in a way that is detrimental?

[0] https://news.ycombinator.com/item?id=45927558

Re: AGI fantasy is a blocker to actual engineering

#373

Post by john01dav »

simonw wrote: post #12

Tip for AI skeptics: skip the data center water usage argument. At this point I think it harms your credibility - numbers like "millions of liters of water annually" (from the linked article) sound scary when presented without context, but if you compare data centers to farmland or even golf courses they're minuscule. Other energy usage figures, air pollution, gas turbines, CO2 emissions etc are fine - but if you com…

Another issue is that you could, in principle, build data centers in places where you don't need to evaporate water to cool them. For example, you could use a closed loop water cooling system and then sink that heat into the air or into a nearby body of water. OVH's datacenter outside Montreal¹ does this, for example. You can also use low carbon energy sources to power the data center (nuclear or hydro are probably the best because their production is reliable and predictable).

Unlike most datacenters, AI datacenters being far away from the user is okay since it takes on the order of seconds to minutes for the code to run and generate a response. So, a few hundred milliseconds of latency is much more tolerable. For this reason, I think that we should pick a small number of ideal locations that have a combination of weather that permits non-sub-ambient cooling and have usable low carbon resources (either hydropower is available and plentiful, or you can build or otherwise access nuclear reactors there), and then put the bulk of this new boom there.

If you pick a place with both population and a cold climate, you could even look into using the data center's waste heat for district heating to get a small new revenue stream and offset some environmental impact.

1: https://www.youtube.com/watch?v=RFzirpvTiOo

Re: AGI fantasy is a blocker to actual engineering

#374

Post by zzzeek »

nasmorn wrote: post #326
Capricorn2481 wrote: post #233

Earlier quoted context omitted.

Okay, but a lot of people are calling environmental and content theft arguments "political" in an attempt to make it sound frivolous. It's fine if you think every non-technical criticism against AI is overblown. I use LLMs, but it's perfectly fine to start from a place of whether it's ethical, or even a net good, to use these in the first place. People saying "ignoring all of those arguments, let's just look at the t…

The environmental argument is frivolous as long as people fly to Vegas for the weekend or drive a F150 to the office. Why is this as special domain?

debating the environmental and public health effects of AI negates the possibility of debating those same things with respect to cars/trucks ?

Re: AGI fantasy is a blocker to actual engineering

#375

Post by tim333 »

Etheryte wrote: post #2

Many big names in the industry have long advocated for the idea that LLM-s are a fundamental dead end. Many have also gone on and started companies to look for a new way forward. However, if you're hip deep in stock options, along with your reputation, you'll hardly want to break the mirage. So here we are.

I figure it's more like steam engines and flight. While steam engines were not suitable for aircraft, experience building them could carry over to internal combustion engines. I imagine something like that with LLMs and AGI.

Re: AGI fantasy is a blocker to actual engineering

#376

Post by naasking »

mysecretaccount wrote: post #300
naasking wrote: post #160

Earlier quoted context omitted.

> To simplify it greatly, an LLM neuron is a single input single output function. A human brain neuron takes in thousands of inputs and produces thousands of outputs This is simply a scaling problem, eg. thousands of single I/O functions can reproduce the behaviour of a function that takes thousands of inputs and produces thousands of outputs. Edit: As for the rest of your argument, it's not so clear cut. An LLM can…

I agree with both of you, but scaling isn't feasible with this paradigm. You could need continent-sized hardware to approximate general intelligence with the current paradigm.

> You could need continent-sized hardware to approximate general intelligence with the current paradigm.

I doubt it, if by "current paradigm" you mean the hardware and general execution model, eg. matrix math. Model improvements from progress in algorithms have been outpacing performance improvements from hardware progress for decades. Even if hardware development stopped today, models will continue improving exponentially.

Re: AGI fantasy is a blocker to actual engineering

#377

Post by killerstorm »

jimbokun wrote: post #361
killerstorm wrote: post #276

Earlier quoted context omitted.

AlphaEvolve demonstrates that Google can build a system which can be trained to do very challenging intelligent tasks (e.g. research-level math). Isn't it just an optimization problem from this point? E.g. now training take a lot of hardware and time. If they make it so efficient that training can happen in matter of minutes and cost only few dollars, won't it satisfy your criterion? I'm not saying AlphaEvolve is "AG…

The "context window" seems to be a fundamental blocker preventing LLMs from replacing a white collar worker without some fundamental break through to solve it.

It's to early to declare something "fundamental blocker" while there's so much ongoing research.

Re: AGI fantasy is a blocker to actual engineering

#378

Post by wonnage »

nasmorn wrote: post #326
Capricorn2481 wrote: post #233

Earlier quoted context omitted.

Okay, but a lot of people are calling environmental and content theft arguments "political" in an attempt to make it sound frivolous. It's fine if you think every non-technical criticism against AI is overblown. I use LLMs, but it's perfectly fine to start from a place of whether it's ethical, or even a net good, to use these in the first place. People saying "ignoring all of those arguments, let's just look at the t…

The environmental argument is frivolous as long as people fly to Vegas for the weekend or drive a F150 to the office. Why is this as special domain?

Driving a massive truck in the city is stupid too and most short flights should be replaced with high speed rail. And AI wastes a monumental amount of resources.

Re: AGI fantasy is a blocker to actual engineering

#379

Post by Faelon »

derektank wrote: post #238
abathur wrote: post #174

Earlier quoted context omitted.

Agriculture feeds people, Simon. It's fair to be critical of how the ag industry uses that water, but a significant fraction of that activity is effectively essential. If you're going to minimize people's concern like this, at least compare it to discretionary uses we could ~live without. The data's about 20 years old, but for example https://www.usga.org/content/dam/usga/pdf/Water%20Resource%2... suggests we were us…

The vast majority of water in agriculture goes to satisfy our taste buds, not nourish our bodies. Feed crops like alalfa consume huge amounts of water in the desert southwest but the desert climate makes it a great place to grow and people have an insatiable demand for cattle products. We could feed the world with far less water consumption if we opted not to eat meat. Instead, we let people make purchasing decisions…

Thanks for the sanity!! I wish more people understood this

Re: AGI fantasy is a blocker to actual engineering

#380

Post by ben_w »

fmbb wrote: post #62
Aransentin wrote: post #50

Earlier quoted context omitted.

Growing almonds uses 1.3 trillion gallons of water annually in California alone. This is more than 4 times more than all data centers in the US combined, counting both cooling and the water used for generating their electricity. What has more utility: Californian almonds, or all IT infrastructure in the US times 4?

Depends on what the datacenters are used for. AI has no utility. Almonds make marzipan.

Marzipan is fun, but useful?

AI is at least as useful as marzipan.

Post reply on HN