and look what playing fair and tying its hand behind the back got the us. Hacked by its opponents and driven towards civil war. The limitations on secret services are part of the fiction of the lawful western world-order.
Imagine if you will- a 3d sintered (https://en.wikipedia.org/wiki/Sintering) sand structure. Placed in it is heat collecting material- aka salty-brine- which stores heat as a kind of battery. Placed within it are waxmotors as actuators. Placed within it are nodes with https://en.wikipedia.org/wiki/Silica_gel - the whole thing is literally dumb as a printed rock, but also capable to do work powered by the sun.
It collects moisture from the air, distills the water out of the silica gel and stores it- then releases it in the evening to plants at the root. A terraforming rock, printed in-situ, with materials available in situ (well except silica gel, optical fibers and wax-motors).
I've also thought this. There are loads of offshore wind farms in the North Sea and I bet the cost per kWh at the offshore substation is a fraction of the cost compared to the same electricity once stepped up for transmission through the undersea cables to the shore, possibly curtailed due to grid congestion and then transmitted still further to the urban load centers.
If only there was a reliable, mature process to turn air\seawater plus energy at the windfarm into a synthetic fuel (ammonia or methanol maybe?) that could be then stored and used to refuel green shipping that burned no bunker fuel or diesel?
Eh, how is this self-inflicted, when you have a radical has-been-empire blocking a global arteries by waging relentless proxy wars. Stop it with the "only white people can be actors on the global stage"-ism, the us basically clashes with Iran once every president (directly or indirectly) - and not by choice.
Well, you can play “who started it” to no end, but technically direct country-to-country assault started with Iran’s missile barrage on Israel.
Now of-course this was preceded by the killing of IRGC members in Syria (and some by beeper casualties in Lebanon), but they were there handling Hezbollah attacks on Israel.
Are you referring to the start of the 12 Day War here? If so, are you leaving out Israel's unprovoked airstrikes on Iran out of ignorance or dishonesty?
I am no fan of Iran but the white people in question literally went there to wage war with the so called has-been-empire and got their butts spanked, despite EVERYBODY warning them for years this would have been an unwinnable war
It does make one wonder how many nations off China's coast are taking notes on how to asymmetrically shut down waterways even against a giant modernish navy. Or Singapore and neighbors... If the world cops can't guarantee safe trade anymore things are going to get strange, and not in a good way.
It was blatantly obvious Iran would do that if attacked. This has been known for decades.
The only thing previously stopping them holding the strait "hostage" was the threat from the US of doing an "Operation Praying Mantis v2" and destroying their navy and attacking them.
And when the US did that anyway, they had nothing to lose.
And Iran attacked. Relentlessly every day up to that war. It does not get into western news much. But in Lebanon. In Yemen. In iraq. In ghaza. There proxies marched and attacked. The west is to weak, to wait for another pearl harbour.
Look at life, or its systemic counterpart capitalism. Look at that attempt at an exponential. In that curve is all the cruelties, all the danger, all the colonialism, by eating other curves, all the failure because physics is linear on a sphere in the sun.
Everytime somebody tries to emotionalize any discourse around any of these topics, look at the defining curve and think for yourself - "yeah, thats life, still preferable to death." - then go and solve that problem, or go an be absolved from that problem. Any vital system is expansive, the trees colonialize the grass by throwing shade. Equality - and thus peace can only be reached by having a equality of arms, by copying working mechanisms.
Nature has existed at a relative equilibrium for billions of years. It changed, but there was never a single thing that could take over the entire system.
It took humans a few hundred years under the current system to start a mass extinction event and begin dismantling the ecosystem we rely on. And we are ready to go 10x further if it means someone somewhere else makes a few more dollars.
Cancer is the most apt comparison, actually. At a system level we don't care for survival or long-term viability, only immediate replication and exploitation.
That is - a anthropomorphization of nature- assuming what is nice and "stable" to me, is not actually furiously attacking one another.
Nature had buildup of warfare materials (the plastic that was trees before fungi digested it).
Nature had campaigns ebb and flow, due to iceages and warmspells, had island holdouts being wiped-out by systems adapted to furious combat on continents (invasive species).
Your picture of nature is very much distorted by a rejection of a rather cthullian reality.
What you mistake for peace is a coldwar equilibrium, where all sites furiously roll the dice of research for new war-tech- and then humans came along. Humans are very much a natural system doing natural things, very successful though.
Cancer is natural, "natural" does not mean good. This focus on things being "natural" is very strange when we're discussing artificial constructs like the economy or technology.
Technology and capitalism are not natural phenomena, they are political ones.
You are conveniently ignoring the comparison to cancer. Capitalism and cancer both demand growth at the cost of everything else. This is what leads to them destroying the systems they inhabit.
Yes, and the focus on growth in general. There is a sustainable amount of goods and services that need to be produced. Capitalism forced the economy to go beyond that many decades ago.
Growth is checked by balancing processes in nature. Every "exponential" in nature is actually a sigmoid. Growth is a means in nature, in shareholder capitalism, it is an end.
Shareholder capitalism explicitly ignores natural constraints in order to chase a constant rate of growth. That's what cancer does, it eats up resources from other cells in the organism in order to do nothing but grow, killing the organism in the process.
So give me a falsifiable point in time, a model you would claim succeeds at the task. One does not get to hotfix-patch updater out of the pressures of reality.
Today is the day.
There is no need to ask. If you want to test SOTA models today, there are obviously only two: GPT-6 Astra and Fable 5.1.
The models listed in the paper are from early 2025 and are no longer relevant, much less on the frontier.
That Claude version is no longer available today, Gemini 2.5 Pro will be shutdown next month, and the OpenAI models are only available via the API today.
Fortunately, a fellow commenter was so kind and did it with Astra. Didn't do that well either [0]. I'm sure GPT-7 will be super mega ASI regardless (since GPT-6 Astra already claimed AGI in the minds of Jen-Hsun, et al.)...
I'll say it till there is any evidence of the contrary, LLMs are not intelligent and their capabilities solely within the realms of well tailored training data. "Just" having been trained on every rule, strategy guide and likely most games of chess on the world wide web isn't even enough for an LLM to play that game reliably. Yet the same model could code a competitive chess engine, just like a model struggling to count can write advanced maths papers. Fascinating tools, but tools nonetheless.
Maybe watch some HuskIRL videos to temper your expectations. Sure, frontier models providers may alter their harnesses to better target chess, but that’s lipstick on a pig imo. The models themselves are not, in isolation, capable of solving general tasks. We haven’t modeled intelligence sufficiently. We’re in a local minimum and throwing billions of dollars at a gamble that that local minimum can facilitate the concentration of wealth even further and fully realize the American dream of eliminating the middle class.
Really? It was being sold as a total replacement for jobs like software engineering and being an attorney, but its looking a lot more that its just going to be a tool those professions use and doesn't actually seem to be taking jobs away.
I find it amusing that you're describing a huge misallocation of capital and a society enabling such, and that is the optimisitic scenario (in my mind anyway).
I very much agree that the next models will be better, heck, I still suck at hobbyist training and could probably coax t5 to do better in Chess specifically, just need to get loads of data from Stockfish.
Thing is, given what GPT-6 Astra was trained on and what models of a similar class can do (including developing a competitive chess engine), it is often paradoxical and somewhat surprising how little these models have gained in actually capability that is in the training data, but not RLHFd to hell, so to speak. Tracking the state of pieces, I suspect given similar in Sudoku [0], is what these models struggle with in game settings, whilst tracking the state of code changes can be reliable over 250k tokens. Essentially, for the latter they were trained in the specific manner that lead them to abstract the capability, but that doesn't track to the former, which is a massive difference between LLMs data focused training and human learning.
So yeah, GPT-7 or any upcoming/present LLM could do massively better in Chess than GPT-6 Astra, but not because the approach was emergent out of pure data. Rather, it requires a very specific training data type and stack for a model to gain capabilities that track a specific task long enough to adhere to the rules of a game such as chess.
I'm wondering if instructing it to track the board state in a file would make a significant difference then.
It reminds me of the ARC-AGI-3 issue where not dropping the thinking tokens between turns or something like that + a new context compaction method increased the performance dramatically. However, I think that is not applicable here.
So what is the supposed leap?
One agent per option to change, evaluating the board state that there move would create, by having a army evaluate the remaining piece options and average over that? Wee-Free-Man as a hierarchical army ?
Pet-LLMs trained on one thing?
Honestly, for intelligence I don't know and I doubt anyone can claim to know. Maybe JEPA, there is potential concerning some shortcomings inherent to LLMs but it has its own, maybe scaling up the electron microscope stuff Google just did (though the connections are inferred), maybe future implementations of autoregressive and diffusion LLMs can at some point address its issues after all, maybe something else entirely.
All I know is, AGI, as in actual intelligence, is quite a massive accomplishment to claim and we shouldn't loose sight of that fact, especially as "not being intelligent" does not make these models any less impressive, fascinating to work on or useful in many tasks. Personally, the only thing I am fairly convinced on is that if we were to find a way to create actual intelligence, it likely wouldn't start out as useful as todays LLMs are and may thus be dismissed early. But again, pure speculation on that front.
If for leap you just mean more utility from LLMs as they are, then I'll pretty confidently put my money on higher quality, not more, training data for a wide range of verifiable tasks. What makes maths, coding, etc. comparatively easy to make gains in (though less verifiable tasks can also make similar as seen with the writing in Kimi K2).
3. Handing a people that can not progress, exponential tech, without which they regress, but which also kills them all when they regress, is a trapdoor function worthy of St.Fermi .
reply