It converts it to co2 and its a structure like carbon capture stuff. Im not big on carbon capture but if you running this thing anyway it might make sense to run the output into some carbon capture scheme as it should reduce both the production and running energy since it can use some of what this is already doing as far as pulling in and exhausting the air. might be good for the exhaust to go down an old well or something to.
Not sure if there is much chance for effective carbon capture. The article states that this works for getting rid of very low concentrations of methane (so burning is not possible). That means that even with the methane 100% turned into carbon, we are talking about very small concentrations.
well there would be the native co2 in the air its taking in too. My point is if it was worth it enough to do on its own its already done most of the heavy lifting so I bet if a carbon capture technique was worth it, it would be riding the output of this.
This cable length is 765km, but in the Canadian province of Manitoba we have a 900 km transmission line built 50 years ago and still running strong for the most part.
AI does have little to do with it, but we can’t do housing the way people want housing. The land does not exist in sufficient quantity, in the desired areas, without other strings attached (such as private ownership). And it would still take a decade to build it all because there aren’t enough tradespeople in the places where you want the housing built.
This is complete and utter bullshit. We have enough of everything to start solving housing this second. Workers aren’t a problem, locations aren’t a problem. We lack the political will to do it, read: we don’t want to do it. Having “AI” tell you why you don’t want to do it is just wasteful
I won’t argue that AI won’t solve the housing problem. And I agree that we can build a bunch of housing. But it won’t be where people want to live, or it won’t be affordable. I’ve got people in my town screaming for affordable housing. Even with subsidies its hard to get things going when the local municipality is practically bending over backwards. Why? Because it has to be on a bus line. It has to be within walking distance of X services. And all the land that fits those criteria is millions of dollars an acre. Even if you could find them, the contractors can’t find enough qualified, reliable workers at premium rates to service their million dollar home builds. I’m in the industry and I don’t care how much “will power” you have; short of taking land through eminent domain and using it for free, you won’t have anyplace that meets any kind of criteria for livability. Hell, I could go buy 1000 acres just an hour down the road for $1M and put up 10,000 houses that only cost $50k each to build. Thing is, nobody is going to buy them. There is literally no demand, even for cheap housing, that takes an hour drive to get anywhere useful - and if you get closer in, you won’t find land that’s affordable. Heck, by the time I extended infrastructure to them or built it out, it would be 3-4 years before the first resident could move in, and that’s with zero delay on any governmental paperwork.
There’s nothing AI could do to help with that. It’s not a technology issue, it’s purely a political will issue. We could house every single homeless person in this country with no problem whatsoever. Right now. Today. But we choose not to.
Maybe AI could solve it – at least, that’s what Scott Alexander has proposed back in 2014. His idea was that of an AGI that would optimize human life (or the universe itself, I guess) for human values instead of profit or other things that drive the whole Moloch problem he thoroughly describes. I imagine housing would also be solved along the way lol
There is this silly idea that SkyNet will turn sentient in one place, then take over the world. That’s fine for a movie, but IRL it’s more likely to work like nukes: once the idea is out there, people from multiple countries will develop their own at the same time.
There won’t be “one” SkyNet… there will be at least a dozen.
“That is a very interesting and speculative question. There is no definitive answer to how many skynets are there going to be, as it depends on many factors, such as the definition of skynet, the level of AI development, the ethical and legal regulations, the human-AI interactions, and the potential scenarios of AI rebellion.
Based on the current evolution of AI systems, it is not a simple question to answer, as it involves many uncertainties and assumptions. However, one possible way to approach this question is to use a probabilistic model, such as the Bayesian network, to estimate the likelihood of different outcomes based on the available evidence and prior knowledge. Alternatively, one could use a creative and imaginative method, such as the scenario planning, to explore various plausible and alternative futures based on the key drivers and uncertainties of AI development.
In conclusion, there is no definitive answer to how many skynets are there going to be, as it is a highly speculative and complex question. However, there are some methods and perspectives that can help us to think about this question and its implications. Ultimately, the future of AI and humanity depends on how we design, use, and govern AI, and how we adapt and evolve with it. 🤖”
You gotta love the bot for adding a cute 🤖 emoji at the end… and on the bright side, it repeated itself, so at least hopefully it doesn’t seem to be here yet.
That’s also assuming true artificial intelligence isn’t just an emergent property of electrical grids, networks and computers. Like that thing where it seems like computers are listening, maybe they are.
It could be… but electric grids and networks are barely on the billions of nodes scale (at best), with a behavior that restricts them as much as possible to a binary “works / doesn’t work” state, organized in topologies designed to stifle any abnormal behavior… while current LLMs are already on the 100 trillions of parameters scale, each simulating a neuron trigger behavior, organized in topologies to maximize the effects of that behavior.
What could get interesting, is getting a billion smartphones with a neural network of a few billion parameters each, all hooked to a network with just some dumb monkeys standing in the way of full integration. People on the Internet already show emergent behaviors they wouldn’t be showing otherwise; it will get interesting when they get manipulated by more and more complex AIs, trained in turn on their own output post-processed by people.
Best case scenario, we’re going towards a tighter integration between humans and machines.
BTW, the premise for the original pre-production script for The Matrix, was that the machines used humans as neural processing nodes; that’s why Neo could gain access to and control the machines, because all humans had the machines’ code inside them, just needed the exploits/bugs to access it. They dumbed it down to “humans are batteries” in the final version, because 25 years ago they thought the audiences wouldn’t get it (and might’ve been right). But now we can see that who’s whose auxiliary neural processor, might change over the next couple decades.
The team named the supercomputer DeepSouth based on IBM’s TrueNorth system, which started the idea of building computers that act like large networks of neurons, and Deep Blue, the first computer to beat a world chess champion.
The name also gives a nod to where the supercomputer is located geographically: Australia, which is situated in the southern hemisphere.
I mean ok, but still, to call anything related to a brain DeepSouth 😶
The concept of the Deep South, a geographical region historically associated with bigotry, injustice, ignorance, poverty, etc., in an American context is simply non-existent in an Australian one. As such, the irony of that name doesn’t really apply outside of the United States.
Interesting, in this particular case it’s implementing a single operation, but I can imagine they can implement other single operation dedicated chips as well. So I’d expect ASICs but no CPUs
By setting the conductivity of each transistor, we can perform analog vector-matrix multiplication in a single step by applying voltages to our processor and measuring the output
Still, i don’t think it’ll need to get much more complex to be very useful for AI workloads.
People have been discovering that more, and simpler, calculations seem to work better? the trend in AI workloads seems to have gone from FP32 -> FP16 -> INT16 -> INT8 and possibly even INT4?
Seems like just having lots of simple calculations is more efficient/effective than more complex stuff.
Well these chips perform analog math, which means high precision high speed. It’s not as accurate as fp32 as in repeatedly and deterministic outputs, but that’s def not a problem for a deep and wide neural network such as used by llm
The lake was a runoff for the colorado river back when farmers over used water and the leftover was dumped ino that “lake”. The lake in its current state is too saline and dried up to ecologically be stable. The buildup of farm chems over the year cause dust in problems in socal when winds picked it up.
Well, when you put it that way using a part of the country we already ruined to try and help us not ruin any more of it, it sounds like a damn good idea
I’ve passed by this ‘sea’. I invite you to visit, if you’ve ever wanted to directly experience what southern California might be like post-humans.
A multitude of enterprising resorts quietly decaying in the harsh inland sun. The only sound is that of hot wind gently peeling paint off so many little abandoned liquor stores and gas stations in between.
Nobody swims. The fish don’t live, and the errant fowl take off just as quickly as they land. This time capsule of the 1960s smells like death, and its strangely yellow dust is concerning even if you didn’t know how much fertilizer runoff was dumped in there by surrounding date farms.
Now I want to visit. I mean, there should be 1000 and 1 reason to visit USA and California in particular, but that’s the first one which really gives a feeling.
You’d also like the 3rd largest city in California, California City. Both are portraits of a generation that could afford to dream big and avoid financial ruin.
Bakersfield too, if you want to find life hanging on a bit better than some of the inland counties. California has a lot of strange feelings if you take the time to look in the right corners.
A desert megacity next to the sea? Los Angeles.
Incredibly posh people surrounded by the destitute? Newport beach in orange county.
A slow-motion wave of impenetrable fog gently cresting the mountains around it? San Francisco.
This state has a lot more to it than TV and movies let on. Don’t just drive it, check it out on motorcycle or bicycle too.
If a person IS strangely into all of that stuff then I actually recommend the annual Fallout New Vegas festival they hold in Goodsprings Nevada, instead.
Fun fact, the beach is made entirely out of barnacles and it smells like someone ate 10 pounds of salmon and then ripped ass straight up your nose. Don’t go in the water, you’ll die!
The entire thing is a lesson in the hubris of man. It was created as a major ecological impact of a failed engineering project. It’s being destroyed by irrigation.
Looks like we have our next Discovery channel reality TV show. Which cast member finds a bomb in today’s episode? Tune in to Lithium Blast at 9 o’clock central to find out!
interestingengineering.com
Active