Since humans can't manage fusion, the US puts millions into AI-powered creation Hot off the heels of the US Department of Energy's (DoE) sort-of nuclear fusion breakthrough, the agency is offering up $33 million for researchers that can wrangle artificial intelligence, machine learning, and other data resources to the cause. The aim is to get machine intelligence to speed the analysis and simulation of …
It's not that Google has amnesia. It's just that history starts from its IPO, so there is nothing older than that to remember in the first place.
SLAC = Stanford Linear Accelerator Center. Maybe Encyclopedia Britannica can be a good start? In any case, with SLAC you are in the right place with slac.standford.edu.
SAIL - Here the Internet may be even more confusing. There are so many places whose founders liked the SAIL acronym that you get lost Googling. Stanford Artificial Intelligence Laboratory goes back to the times of John McCarthy. A seemingly separate page provides a <cough>highly personal</cough> overview of history that, among other things, claims that just about every bit of technology one has ever heard of (including, khem, Google) originated from there. Some of it is very likely true, and at the very least it is entertaining.
Damn, I am old.
Damn, I am old.
Yet your intact memory has info that 'teh intarwehs' apparently does NOT have easy access to.
Robot overlords have a LONG way yet to go!
After reading the article and reference to'fuel pellet', looks like they are trying to use inertial confinement. I do not think that design has much to offer for a sustained power producing reaction as it is the equivalent of setting off a tiny nuclear bomb with lasers.
There is a BBC documentary, out there on ''teh intarwebs', about the pistol shrimp, which uses its claw to create an underwater shock wave to stun its prey. Slowed down it shows a large cavitation bubble collapsing, and also a brief but intense flash of light. i keep thinking that it may be briefly creating a small fusion reaction, which would greatly amplify the effect of the shock wave created by the shrimp's claw. Result in my head looks like a million pistol shrimp claws rapidly repeat-firing to boil water. Eh, just a thought.
The sonoluminescence from the cavitation elicited by the pistol shrimp does indeed include the creation of a plasma, and fusion would not be absolutely out of the question, but the probability would be vanishingly low for even one fusion event, given the "triple product" we could infer from the conditions and the "fuel" species present.
Nevertheless, the pistol shrimp is the inspiration for First Light Fusion in Oxford. Their approach is all about focusing and amplifying shockwaves initiated by a physical projectile.
This post has been deleted by its author
Oppenheimer and others had concerns that the nuclear *fission* bombs might cause a propagating chain *fusion* reaction (of hydrogen and nitrogen in the atmosphere and hydrogen in the oceans). Oppenheimer tasked Hans Bethe and others to calculate if this was actually possible. They concluded it was not. Indeed no A-bomb or H-bomb ever ignited the atmosphere.
If it was that easy, without confinement, we would have had fusion power stations decades ago.
The plasma in a tokamak is really not very dense. There are just a few grams of deuterium and tritium in there at any one time. Magnetic confinement is required to try to keep the plasma from melting or vaporising tiny holes in the first wall of the vacuum vessel, and to try to squeeze the plasma more tightly to increase the fusion rate. It is not necessary to prevent a chain reaction. Controlled fusion very rapidly fizzles if the conditions drift even slightly away from optimum.
You really don't need to worry about a Tokamak, or similar, turning into a nuclear bomb.
Oh! And the pistol shrimp use a poorly understood phenomenon called cavitation. Cavitation is, by the estimate of some who study it, also a possible future source of energy production, since it is suggested that the "crack" of a cavitation event releases more energy than it consumes. That one fascinates the hell out of me and I really encourage the uninitiated nerds among you to check that topic out.
Do you have any reliable sources / references for this thermodynamics defying trick that you allude to regarding cavitation? There have been a lot of "woo" / fraudulent claims in that sphere over the years. Although there are (flawed) patents going back to the 70s claiming to be able to use it to enable fusion break even, I don't think I have seen anything convincing about cavitation in and of itself being an energy source, merely an energy concentrator or possible route to more efficient steam generation. Thanks in advance.
Pistol Shrimps are fascinating, but I doubt they break physics. Also curious now about sonoluminescence. There's a similar effect when a bullet passes through a ballistic gel block. If it has the right combination of velocity and cavity generation, there's a flash of plasma when the gel snaps back. AFAIK that's just a compression effect, and a similar collapsing cavity trigger that isn't sound related.
"Henry VIII", "Hammurabi", and "Paleozoic" pages were created by historians and geologists (maybe with paleontologists), not AI experts who would be well advised to create more pages about the history of their own field. John McCarthy and SAIL are positively Paleozoic, maybe specifically Cambrian (in the explosive sense) in the context. Try googling SAIL and count the organizations that you find before you get to Stanford AI Lab. I was shocked - and I knew what I was looking for.
FWIW, the phrase "Silicon Valley thinks that history starts with Google's IPO" (or something pretty close) was coined by an eminent (Stanford) historian. I'll leave googling it up to you.
No disrespect, but if you knew what you were looking for then your google-fu needs a workout. Google search did used to be much more powerful and targetable than it is now, sadly, but you can still reduce the greediness of its matching by using quotation marks. Search for "Sail" "Stanford", for example.
Last time I checked (10 years ago (ish)), SAIL's original DECtape "Permanent Files" from 1966 to 1972 and then the 7-bit DART tape archives (Dump And Restore Technique ... essentially full system backups of the SAIL PDP-10), from 1972 to 1990, were available to researchers at Stanford's Green Library (in "The Digital Collection"). Access is (was? see below) restricted to people who have permission from the original authors.
I know there was some effort to put all that into a more modern archive format, and then put the results online, right around the turn of the century. I do not know how far the effort managed to get.
Note that the interaction between SAIL and SLAC (and The Big Dish folks) mostly wasn't official, but tapes between the three were exchanged fairly regularly. SAIL mostly helped with knotty algorithm wrestling on large (for the time!) data sets, robotics, and imaging..
The following map shows how close these three campuses are to each other, with the main Stanford Campus upper right. The now sadly demolished D.C. Power[0] building, where SAIL was located, is currently home to Portola Pastures, a horse establishment (bottom center).
https://www.google.com/maps/@37.4071914,-122.1815786,4723m/data=!3m1!1e3
[0] Nothing to do with Electricity ... it was named after Donald C. Power, a corporate director of GT&E, who had built the building as a research lab, and then gifted it to Stanford. Lore has it that SAIL (originally SAIP, "Project") got the last part of it's name from the building ...
Indeed - the recent "breakthrough" was achieved after they tried AI to calculate a new geometry for the 192 lasers.
What everybody seems to have ignored about that gain of about enough energy to boil 3 kettles is the single use fuel capsules are hand made in gold and cost about $10,000 each.
"What everybody seems to have ignored about that gain of about enough energy to boil 3 kettles is the single use fuel capsules are hand made in gold and cost about $10,000 each."
True, but then the early internal combustion engines were hand made, didn't work all that well and only the very, very rich could afford to build a car around them. Fusion isn't even at that stage yet, but cars are affordable for many, many people these days. If and/or when sustained fusion reactions become possible, the infrastructure for making the fuel pellets or whatever method is used then will become a mass production problem, not a hand crafted, one at a time operation.
True but a poor analogy, with mass production cars went from costing a thousand or so dollars to a hundred or so, a reduction in price of around one order of magnitude. For inertial confinement fusion to be viable the cost of the fuel pellets would need to be under 10¢ each - a cost reduction of over 5 orders of magnitude which is a bit of a tall order.
Also the system used is able to fire about once a day, for viable fusion power generation they'd need at least 100 per second.
All in all it was an interesting development but not a significant advance towards viable power generation with fusion. Magnetic confinement still seems to be the more scalable option.
> True but a poor analogy
Relatively speaking only; Internal combustion engines did go through some quite surprising iterations and were initially ridiculously inefficient too (19th century Lenoir automobiles were 2500 cc / 1.5 hp!). It took many decades before the Otto/Diesel cycles were invented and then perfected to become the mean & lean powertrains we know today. Getting from the primitive Lenoir engines to Diesel engines took almost half a century.
My point is that fusion is still at that initial point, it still needs a Nikolaus August Otto and a Rudolf Diesel to get commercially efficient. But since everyone builds on what his predecessor has already invented and explored (there would be no Diesel without Otto, and no Otto without Lenoir), each step forward is a good one.
This is all true but still misses the magnitude of the problem, from 2500cc giving 1.5hp to 2500cc giving 150hp took almost 100 years and is only a gain of 2 orders of magnitude.
For fusion a gain of 5 orders of magnitude is required and on the duty cycle side an improvement of 7 orders of magnitude is necessary. (rather than firing once a day, you would need it to fire 10million times a day to be useful).
Then there is the efficiency of the reaction, only a miniscule fraction of the fuel in the pellet fused, they need most of it to fuse to get around 100 times as much gain to make it workable.
Not saying this isn't interesting just that nobody should think this will translate to a useful fusion generation system any time soon.
While I fully agree that this method is not likely to be the solution to cheap fusion power, taking the analogy to the extreme, we've just reached the stage of the first artillery cannon, ie a controlled explosion in a cylinder, a long way from the 1000's per minute needed to make a car work, let alone the fuel feed methods and chemical energy translation into motion :-)
On the other hand, other methods are being tried and there's a lot more money and brain power going into this research :-)
> (rather than firing once a day, you would need it to fire 10million times a day to be useful)
I think we both agree that this laser-pellet procedure isn't likely to be the one which might one day power our cities.
But you need to keep in mind that there have been quite some UFOs among the first internal combustion engines too, like for instance the (patented!) "Pyreolophore" which around 1900 used controlled dust explosions!... Somebody around that time would had been justified to doubt this new invention would be able to literally move the world, a little more than a century later. And indeed, the Pyreolophore system was clearly not the right way to do internal combustion engines.
Any new technology starts with an idea, like "energy of confined explosions" or "energy of hydrogen fusion", but at that point nobody has the slightest clue about the "how", so inventions tend to explore all kind of strange dead-end ideas, until one (or a couple) eventually show their potential and get adopted and perfected. Fusion is still in this "we have no clue how to do this, so we keep exploring random ideas looking for the right one" phase. And unfortunately, unlike internal combustion engines which any mechanic could build and experiment with in his garden shed, fusion requires eye-wateringly huge investments, which is why it takes a lot longer to get out of the trial & error phase.
I don't hold my breath, but I'm optimistic it will eventually happen. If only because there is big money in being the first who gets this working. Let them keep throwing things at the wall and see what sticks.
The problems for early internal combustion engines were twofold, first getting the fuel/air mixture right and secondly increasing the efficiency of the engine both of which were really just questions of engineering to perfect carburettors (and later fuel injection) and as for efficiency the principles of heat engines were laid out by Carnot in 1824.
For inertial confinement fusion the problems are 1) cost of fuel delivery 2) speed of delivery 3) efficiency of reaction.
Mass production might be able to come up with something like cine film with a blisters of fuel along it so it can be reel loaded which will also solve the speed of delivery problem, that's all just engineering and should be solvable with current knowledge, the big big problem is efficiency of the reaction, we don't have established principles for a fusion "engine" like the internal combustion pioneers had instead we need new science which might happen tomorrow or maybe not for 50 years.
I agree fully we'll get to sustained fusion at some point, I also doubt if it'll be soon and I have some doubts if it will be a cost effective form of energy production.
I agree fully we'll get to sustained fusion at some point, I also doubt if it'll be soon and I have some doubts if it will be a cost effective form of energy production.
I think there are still.. a few obstacles to overcome. There's a neat video here-
https://www.youtube.com/watch?v=G9uJYk-8w2s&t=36s
Where Kyle Hill goes over the announcement and some of those challenges. Also covers the LLNL's LIF(F)E Engine, which will probably scare anti-nuclear types, especially as the graphic mentions weapons grade Pu. But the neat trick is by using a hybrid fusion-fission process, it allows for potentially cheaper fuel. So aneutronic fusion is nice, because it doesn't generate the fast neutrons that are harder to shield. Instead, it relies on neutrons to create fission heat in a 'blanket'. That looks to have a few benefits, so first layer is lithium, which produces tritium when it's zapped by neutrons. Then wrap that in spent fuel, depleted or recycled plutonium, which are good neutron absorbers, so heat up and don't/shouldn't go critical.
Then when all that works, the gain's shown as 120-600x and it becomes a kind of tritium breeder reactor, so lowers the fuel cost and becomes more sustainable. But that's all a little ways off, but has far, far higher potential than this garbage-
https://gridwatch.co.uk/WIND
Because it doesn't rely on the weather. Just political will to fund future tech instead of making bigger 1,000yr old, fundamentally obsolete tech.
> I also doubt if it'll be soon and I have some doubts if it will be a cost effective form of energy production.
"Cost effectiveness" is in the eye of the accountant. Fusion is first and foremost an energy requiring a cheap(er) and (more) readily available fuel: Hydrogen is the most abundant element in the universe and can be found everywhere, so there will be a lot of cases where it's simply the only possible solution (you won't be sending coal to Mars for instance).
Common Hydrogen is useless as a fusion fuel. Deuterium and Tritium are required neither of which are tremendously common and both are expensive to extract and purify, especially Tritium.
Sea water has about 0.0156% Deuterium which is considering the volume of the oceans is a fair amount but there are only 1 tritium atom per 10¹⁸ hydrogen atoms which is almost insignificant. Tritium can be generated by bombarding Lithium with the neutrons from the fusion reaction but that's very difficult as ways of doing this in anything beyond a lab experiment have not been developed. Also standard off-the-shelf Lithium does not work, the reaction needs Lithium⁶ which is a controlled substance refined in tiny quantities for use in thermonuclear bombs.
Another potential fuel is Helium³ which consists 7.2 parts per trillion of the atmosphere but is quite abundant on the moon so that could be a future fuel.
So the fuel for fusion is neither cheap or abundant and the cost of building a workable fusion reactor combined with the running costs may possibly be prohibitive when compared with the alternatives. Also fusion while clean suddenly gets a lot less clean if you want to generate Tritium from the neutron/Lithium reaction as that leaves some nasty radioactive residues.
This post has been deleted by its author
They say nuclear fusion, we say NEUKlearer HyperRadioProACTive IT is the Future for Unlimited Power and AWEsome Energy.
They say another Manhattan type Project, A.N.Others say of an AWEsome Weaponised Experiment .. Oh I see. Let the Greater IntelAIgent Game[s] begin ..... and may the very best of novel beings who dare care share win win.
Merry Xmas, El Regers. Have an nice one and prepare yourselves for an absolutely fantastic and frantic New Year, for there be universal secrets to spill and tall tales to tell and honestly prove fundamentally true.
Pattern matching. Yeah, sure. You usually need to tell it about some of the patterns and do classifications and all of that stuff. So... Yeah, but nah.
I myself have greater hopes for tocamacs or maybe even stellarators (however they may be spelled, keep all typos you find, merry Christmas), though I think the latter are no longer en vogue.
> Pattern matching. Yeah, sure. You usually need to tell it about some of the patterns and do classifications and all of that stuff.
And why do you suppose that's not feasible? Researchers may have a good idea what sorts of features they're looking for - just not how to spot them in a maze of data.
Up until someone makes an AI that out performs politicians:-) .... LateAgain
That faux hurdle was long ago dispatched to the dustbin of history and has more than just politicians and Google realising they are easily bettered by that which over which they have no practical physical command and control ..... Alphabet/Google reshuffles to meet ChatGPT threat and Sundar's not having a happy holiday ..... other than trying to pull the plug on a Golden Goose of an AIDevelopment which will just transfer itself to foreign shores elsewhere to aid and abet that which and those who many would squeal are alien competition and deadly market share rivals.
The madness that abounds and compounds the enigmatic conundrum is that the bettered do not recognise their obvious dilemma, which would be quite sad if it were not so tragic and well deserved.
I recently realised that AI researchers are so confident that AI works that they only ever produce AI systems to do things with, to, or for OTHER PEOPLE. I'd be very interested in a reference to an AI researcher who produced a system to take important decisions for or about her/him.
Apart from that, I was wondering whether to bet we'd get Fusion before or after we got actually useful and trustworthy AI systems, but I guess that if fusion relies on AI, I'll not be holding my breath for either.
I'll get my coat, it's the one with a copy of 'I Robot' in the pocket.
The best Top Secret Secure Stealthy AI Systems are useful and trustworthy, it is the negative reactive human response to ITs Solutions for established hierarchies and oligarchies which are the commonplace problem persistently proving to itself that humanity itself is too easily always able to be rendered extraordinarily flakey and too crazy to fully comprehend and exercise AI in Absolute Command and Control with IT and Media Leverage ....... and thus has no viable effective leadership part to play in Future Greater IntelAIgent Games Plays.
How do you want to cook those Fabergé type Easter eggs?
On the one hand, oodles of data and computers are a match (usually) made in heaven.
On the other hand, modern "AI" (hah!) is all GANs, which are good at producing outputs that perform as well as the best inputs, only in a weird way that is totally lacking in explanatory detail.
So we can spend a lot of money to create lots of new designs for fusion systems that could break-even but don't quite manage to generate any usable power, whilst being terribly, terribly complicated and shiny piles of machinery that no human can make sense of. The Uncanny Valley of particle physics.
is to find a way to maintain a stable reaction. It's like riding a bike. As long as you can do it you don't really care how you're doing it or about the physics that make it possible. (Though in the case of fusion the physics aren't exactly a mystery.) It's an engineering problem to be solved by whatever tools you have to hand.
If you take a look behind the veils of the virtual curtain, this is one major part of the bigger picture to see there .... and ponder on what is next to be supplied and delivered. And to ignore or attempt to deny it, is a negative reactivity which renders one in realisable physical peril.
amanfromMars [2212250533] ..... shares some pertinent news on https://www.nationaldefensemagazine.org/articles/2022/12/23/pentagon-paves-road-map-for-zero-trust-by-2027[Thank you. Your comment will be displayed soon after reviewing.]
That in a Defense Department is an admirable and laudable aim with a high degree of encouraging success likely possible, however, in any Defense Department turned Rabid Attack Dog will that be a debilitating and ruinously costly in blood and treasure Sisyphean task on a rocky road bombarded with giant boulders and littered with concealed explosive munitions because .... the fundamental and exceptionally spectacular element of unexpected surprise and early prime intervention with effective novel operations will be lost to systemic delays in the new hardened battle communications lines interconnecting siloed networks.
And what do you think of the chances of systems administrations in such a closed reinforced network both recognising and accepting outside foreign paramilitary type alien help which would/may be being offered in those warring situations being readily available and desperately needed?
Will that machine choice depend upon a human decision ..... or is that already an automated artificial intelligence operation with SMARTR* Future AIMachinery nowadays smart enough to ensure humans always fail in their choice exploring any wrong direction?
* ...... SMARTR Mentoring Analysis Reporting Titanic Research
Good advice whenever such is honestly true and also coincidentally quite surreal? Deal with it and make great deals with it for you will not be thanked and feted for anything else lesser rendering one perfectly fit to be pilloried in stocks and at the mercy of raging ravenous mobs.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
