Artificial Intelligence (Previously "Chat GPT")
-
@Kirwan said in Artificial Intelligence (Previously "Chat GPT"):
Using AI to understand;
Google DeepMind, with Brown, NYU, and Stanford collaborators, developed enhanced Physics-Informed Neural Networks (PINNs) using second-order optimizers to uncover unstable singularities ("blow-ups") in fluid equations like Euler, Navier-Stokes, IPM, and Boussinesq. Discoveries include novel singularity families, a λ-instability order pattern in IPM/Boussinesq, and ultra-precise vorticity visuals (Earth-diameter accuracy).
These singularities probe fluid limits, tying directly to the unsolved $1M Millennium Navier-Stokes problem on turbulence and smoothness. The AI-math hybrid enables rigorous computer-assisted proofs, accelerating breakthroughs in physics (e.g., turbulence modeling) and engineering (e.g., aerodynamics).
Real-world benefits include optimized aircraft/car designs for fuel efficiency, improved weather forecasting via better atmospheric models, enhanced biomedical simulations of blood flow to aid cardiovascular treatments, more accurate ocean current predictions for climate and shipping, efficient petroleum extraction, and pollution dispersion modeling for environmental protection—ultimately enabling safer, greener technologies where traditional math falls short.
Yeah, I've been spending a lot of time thinking on this recently also...
-
-
I'm finding I'm increasingly using AI in homelife. I've mentioned medical advice before but just reinforcing that my infant had some health issues and plugging symptoms and blood results into grok gave me real peace of mind as to what the issues could be, and how likely they were. Groks diagnosis turned out to be correct once we eventually went through the system to specialists and more blood tests. I'd never say its a replacement for doctors as such but unlike googling symptoms which generally results in "you and your family have cancer and are going to die" with AI you get much better results in a format that is better than a doctors message at 4pm on a Friday saying you need to get bloodtest done in order for an urgent referral to a specialist early the following week -naturally having us freaked out.
The other area I'm finding it terrific is resolving tech issues. I know enough about tech to fiddle and get myself into trouble. I can generally nut a few things out to get a resolution but sometimes something so illogical happening (like intermittent issues) throws me completely. Just over the weekend I set up 2 wireless access points using old ISP routers to enable better comms to our garage and outdoor entertainment area. A year ago I got them kinda going, like the garage door would work 75% of the time and music streaming would skip every 30 seconds if it would connect at all. Grok was able to isolate the issue being with specific isp routers which have certain settings turned off or unavailable with crazy results being that 'some' websites worked while others didn't. Anyway I was able to bounce ideas off grok and worked out some obscure settings to change which seemed to have done the job.
-
I saw a report this morning saying that the Meta data center in Wyoming will use more electricity than the entire state's houses do combined. That's crazy.
Local governments subsidising this 'gold rush' at the expense of local households is insane, the profit extraction is apparently 10:1 and very few jobs are extracted from them, so who is getting the gold?
I like the possibilities of AI but it is hardly being used for really impactful usage.
-
@gt12 said in Artificial Intelligence (Previously "Chat GPT"):
I saw a report this morning saying that the Meta data center in Wyoming will use more electricity than the entire state's houses do combined. That's crazy.
Headlines are definitely on the 'big scary thing' side. AI uses a lot of electricity. However:
A 2012 report on Wyoming's energy consumption by the end-use sector showed that the residential sector used about 4.6% of the state's total generated electricity.
Edit: Of course this was AI, so the electricty may be used to hallucinate
-
Even if we accept that data, That would still add 5% from one data center.
Fuck it, I checked it - 10% from residential, so this one data center adds that again.
If that cost isn’t being offset for customers (and apparently hasn’t been in other places such as Arizona), its mad that regular working class stuffs will be subsidizing AI slop.
I read that in Virginia data centers consume 40% of electricity
-
We will have no choice but to use nuclear power for that, the power consumption needs just seem too great for any other source to provide.
-
One of my main concerns with AI right now is video, it's still easy enough to pick what is AI but it's getting harder. In 3 - 5 years time it's going to be impossible to tell for Joe Public I think.
-
@nzzp said in Artificial Intelligence (Previously "Chat GPT"):
@Rembrandt if you're in NZ, HealthLine is superb. Nurses on call who are pragmatic and will help manage symptoms and risk - and it's 24/7 and free. Really good to confirm you don't need to rush to hospital at 2am for your child
I'm in Oz and there is an equivalent which we used with our first child but my experience with them was 'go to emergency' was their default response over issues that now we're a bit more experienced seems overly cautious. Good to know NZ are a bit better.
Another use I've played with is 'Therapist' mode on Grok. Had a few complicated and extremely annoying inlaw/family issues hit us in a short space of time when we had very little time to manage such issues. Rather than tell some folk to 'go and fuck themselves' I fed the situation into grok as a test and its proven pretty awesome for getting your thoughts straight with solid ideas for de-escalating situations and working towards solutions. Not had any solutions as such yet but at least I haven't made situations 10x worse which is a bit of a win.
-
@No-Quarter said in Artificial Intelligence (Previously "Chat GPT"):
One of my main concerns with AI right now is video, it's still easy enough to pick what is AI but it's getting harder. In 3 - 5 years time it's going to be impossible to tell for Joe Public I think.
Grok is planning an AI detection feature, as there will be things to check for.
I’ll be extremely disappointed if it’s not called Bladerunner
-
As for the power, there is going to be a massive rush to build out power infrastructure . Google and Microsoft have bought half the future output of prototype FUSION reactors.
Fission is being talked about seriously again, with all sorts of new designs.
Bezos was talking about data centers in space.
There is a potential future with all this investment where energy becomes too cheap to meter (and we swap to paying for tokens instead of electricity).
In a more pragmatic scenario I’ve seen proposals to have a two tiered electric payment structure. Cheaper for residential and industry, astronomically more expensive for the Token Factories.
Unfathomable investment is being made to generate cheap electricity.
-
@gt12 said in Artificial Intelligence (Previously "Chat GPT"):
I saw a report this morning saying that the Meta data center in Wyoming will use more electricity than the entire state's houses do combined. That's crazy.
Local governments subsidising this 'gold rush' at the expense of local households is insane, the profit extraction is apparently 10:1 and very few jobs are extracted from them, so who is getting the gold?
I like the possibilities of AI but it is hardly being used for really impactful usage.
From my perspective he one great outcome is practically everyone is suddenly a fan of nuclear again.
-
Is anyone here using ai for engineering projects .. ie I need to design a 60’ aluminum gangway. I’m not a structural engineer, and using ai to design it was incredibly easy and awesome. My problem is: how do I know if it’s actually correct? Designing for 40 psf live load.
-
@W32 said in Artificial Intelligence (Previously "Chat GPT"):
problem is: how do I know if it’s actually correct
Build it and load test it?
Joking, but not 60ft if there's consequences of failure then get an engineer or experienced person involved.
Do not trust AI without a human in the loop
-
@W32 said in Artificial Intelligence (Previously "Chat GPT"):
Is anyone here using ai for engineering projects .. ie I need to design a 60’ aluminum gangway. I’m not a structural engineer, and using ai to design it was incredibly easy and awesome. My problem is: how do I know if it’s actually correct? Designing for 40 psf live load.
Grok in expert mode checks sources and will get you to 80%. The last 20% as @nzzp said needs the human in the loop.
It's not AGI yet, and still imagines facts. For example, some of the models are more than happy to delete tests so you have no failing tests when working with software.
-
To highlight the issue with AI video. This video appears to be AI generated, but 99% of the comments have absolutely no idea, just praising the boy etc. Only a couple of people pointing out that it'll be AI.
Even if it's possible to prove in court what is AI and what isn't (I think that'll be a challenge in itself), the reputational damage to people could be huge. We have all seen tweets that are wrong go viral, and the correction tweet gets barely any attention.
I am at the point where I am just questioning every video I see on X now, which sucks.
-
@No-Quarter and thats X which has community notes which is reasonably good at correcting things. Other platforms are even worse, had a look on facebook the other day, ai news sources (I assume ai) now make up complete fabrications "Paddy Pimblett refuses to fight on UFC gay pride event" and there is zero recourse for correction. Probably still better than their false fact-checking tool..but its a scary state of affairs.
I also saw an old acquaintance posting publicy about Palestine on FB..he has never been even slightly interested in politics, like not even vaguely on the radar.. but I wonder if these algorithms are drawing people in and radicalising them. Add AI video and god knows where things will be in a couple years.
-
Community Notes works pretty well, that's why Zuck is copying it for his platforms. The crowd source approach is a good interim step until we get AI watermarks somehow. XAI have a plan for that, will see how effective it ends up being.
Will definitely need some sort of blockchain/watermark that can't be spoofed for video from politicans and news reports sooner rather than later.
Even I get caught by AI Video now, and I'm looking for it. Good videos are indistinguishable from real videos now.
New antibiotic targets IBD — and AI predicted how it would work before scientists could prove it - Faculty of Health Sciences
