Look, if you continue down that road to GCC countries engaging Iran, then you have a multiparty nuclear armed conflict with combatants stretching from Europe to the Chinese border.
At that point it really does sound like ww3 started from the same causes as ww1 - nobody will win, nobody will know why they are fighting, and most of the fighting will be drones being slung over trenches.
> At that point it really does sound like ww3 started from the same causes as ww1 - nobody will win, nobody will no why they are fighting, and most of the fighting will be drones being slung over trenches.
Name me one war of aggression that ended up being a long term win for the aggressor.
Depends on what you mean by "win". It would be possible to go in, topple the regime and secure the nuclear material. But only at astronomical cost and years of blowback
"Regime Change" has become a modern term for vassalization. We should not be surprised that countries with no reason to be a US vassal, and no long-term ties to the US refuse to remain vassals.
So then what would we achieve? nuclear material is cheap (10s of billions) relative to a multi-decade occupation (single digit trillions). It's undoubtedly true that Iran would revert to it's preferred form of government, geopolitical orientation, and nuclear capability once the US left.
Winning a war means achieving your political goals while preventing the enemy from achieving theirs. Most of the time, you've won the war when the enemy effectively admits they lost.
The lack of will to use sufficient force to win a war is fundamentally no different from not having that force in the first place. Both are equally real constraints on your ability to win the war.
How’d that plan work out in Iraq or Afghanistan, both much smaller, less armed countries? Decades and trillions spent, and what exactly did the US “win”?
Straights have been impossible to force since Churchill tried to force the Bosphorus in 1915. Placing ships in a narrow target area that can be pre-sighted is a losing proposition, a single artillery gun could mission-kill a destroyer in hormuz - mines/torpedos/drones could sink a ship in a place where rescue may not be possible.
Computer science research tends to look like this. Take a seemingly ambitious idea, then spend eons making a version of that idea which doesn't scale and probably doesn't work. But no one is really sure how far this incomplete idea will go. After demonstrating it we realize what the limits and next steps are.
They are saying that models should be invariant to data's orientation - and only sensitive to the distance between vectors. This has a pretty significant effect on reducing the set of possible models, and may stabilize the optimization.
In simple terms, large ML models like LLMs often learn trivial rules such as "if the 21st decimal place of the 5th dimension in the embedding vector is 5 - then the image is of a cat." Learning such a memorization function is usually not what we are trying to do, and there are a variety of techniques to avoid these trivial solutions and "smooth" the optimization geometry.
I don't see how drones don't make all conflicts into WW1. 100 Billion dollars buys about 3.3 million Shaheds assuming the manufacturing is not made more efficient. There are many questions on whether its possible to spend 100 billion dollars on Shaheds, or launch all of them. But this is more than enough to destroy any logistics and transportation infrastructure necessary for a ground invasion.
There are many many countries who can afford 100 billion dollars for stored military equipment that has a long shelf life. The US makes ~50k artillery shells a month at a cost of about 10k per shell.
From my extremely uneducated point of view it seems like that is true and probably what is already happening in Ukraine. However, at some point robots might be able to take and hold ground, and maybe they can be designed to require only decentralized, automated infrastructure to operate that is hard to strike economically even with drones. At that point, may the side with the most robots win.
Of course, once loitering, intelligent munitions make it too dangerous to be an economically valuable human outside of a bunker, we'll need robots running the robot factories, then we get Philip K. Dick's scenario in The Second Variety.
Production is not a hard problem. Iran, a heavily sanctioned country, already has drone production in other countries. That's assuming no other country would want to sell them their own drones to boost their domestic industry, like Turkiye has been doing for Ukraine.
Most of the Iranian drones are quite sophisticated for what they need to do. On a pinch they could replace many of the non-critical components for cheaper parts. They don't need composite materials if they were simply trying to outproduce. Meaning their production facilities could be much simpler than they are currently and still sustain enough output to matter.
Now imagine for the same $10k cost making a cruise missile, instead. This is close to what a Shahed is -- the estimate is $20k-$50k / unit, so close enough.
This is bonkers. Countries can now afford for the same cost * to make not a 10-20 mile range artillery shell, but a 1500 mile effective range cruise missile.
* Defense costs are "fake" to a large degree. A lot of that is really corruption with money flowing from the taxpayers to the arms manufacturers, but still if we go by the numbers...
> Adam Smith pointed out the first pencil costs thousands of dollars, but the second is mostly free. Same dynamic here, but multipled by a thousand.
The shells are already made by the 10 and 100s of thousands, Shaheds are also not a research project, so either one is in amortized serial production now.
What I meant is that a $10k shell doesn't cost that much. Russians are making the equivalent artillery shells for an _order_ of magnitude less for around $1k. A lot of defense costs are just overinflated simply because they can be. The government is spending taxpayer money, it's not really coming from the politicians' pockets. If the kickbacks are just right, they may in fact flow back into the politicians pockets.
A lot of defense spending revolves around overall manufacturing capacity. Deals contain options that won't be executed unless it's war time. These options increase the cost of the deal as the manufacturer needs to keep capacity.
It is vastly more complicated to find targets at 1500 miles than at 20. So drones are effective at destroying big stationary civilian infrastructure and much less at long distance strikes at military targets. Russia's inability to destroy Ukrainian aviation is a good example.
But then with solar and batteries civilian infrastructure becomes much more resilient against drone strikes.
At a certain distance, I'd contend all infrastructure is big and static. Our energy comes from large facilities, without these facilities continent scale infrastructure will grind to a halt at 1500 miles. Rail, power lines, warehouses, factories and trucks are all relatively static. It's not unreasonable to expend a Shahed type drone on a simple semi-truck parked overnight from nearly a continent away. There are only 3 million semi-trucks in the entire US, and I'd be shocked if the country could run without them.
Ukraine tried to come up with drones that can fly over 1000 miles. But drones the size of Shaheds just cannot fly that distance without significantly reducing the warhead. To attack things beyond that range Ukraine have used essentially Cessna. Which is much more expensive and visible on radars.
Instead Ukraine came up with an idea of mass producing extremely simple cruise missiles that could fly 2000 miles and deliver up to a ton of explosives with a cost of 100K and make 1000 of them per month. But then it seems Russia was able to discover the production sites and destroy them.
> It is vastly more complicated to find targets at 1500 miles than at 20.
It's true but they are so cheap that launching a whole bunch and/or improving them incrementally is possible. Yeah they are for stationary targets mostly, for sure. And of course their sounds and relatively low speed does make them somewhat easier to shoot down with short range AA guns and can have automated acoustic early warning system (it's like a flying lawnmower or chainsaw).
1500 mile range is questionable in practice I've read - drones require remote control for maximal value and that's a capability that may not extend nearly as far as the paper range of the drones
They can’t be used for moving targets but for infrastructure they can be effective. At the cost of only a few artillery shells send 10 and maybe 3 will hit.
Another advantage is because of simplicity and cost it allows quick iteration and adaptability. Use honeycomb patterns to lower radar signatures, use specialized antijamming gps/glonass antennas. Engine is too slow? Add a small turbojet. Color too light and visible at night? Paint it gray, etc. That can happen at the speed of weeks and months. Try doing that with Tomahawks, artillery pieces or HIMARS.
The invisible Hand may be monetary policy. The median household may simply no longer be able to afford the median home due to the continued wealth distribution shift brought on by interest rate targeting.
It also forces you to focus on some extremely narrow problem definition. Often, these narrow problem definitions turn out to be features of existing platforms or in the ai age - artifacts of the current model generation.
Anecdotally, the concern I hear from many is that the current positioning of AI as labor replacement doesn't benefit them at all. An expensive AI which simply takes your job or forces you to work harder is categorically worse for people's quality of life.
What consumer benefits is ai driving? at least with industrial automation consumers benefited from new technologies, cheaper goods, and new job categories.
In case someone at Anthropic reads this.. if you find some way to make software developer salaries go up as a result of using your tools, or find some way to fast forward society to that stage of the effect of AI, you’ll have a lot of fans, and even faster adoption.
It would be great if there was some internal “make this benefit Main Street and knowledge workers” department, helping find ways for workers or creators to capture the value of some of the increased productivity.
> It would be great if there was some internal “make this benefit Main Street and knowledge workers” department, helping find ways for workers or creators to capture the value of some of the increased productivity.
If they wanted to do this, they could put their models in a public trust for the public's access and benefit in research, education, etc. Then it could be licensed, pay a dividend like a sovereign wealth fund, etc.
Considering that they copy and train on the sum total of all human creativity, a public trust is something that would be in line with both the spirit, and first and fourth considerations, of fair use doctrine:
1. the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;
2. the nature of the copyrighted work;
3. the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
4. the effect of the use upon the potential market for or value of the copyrighted work.
That way everyone is rewarded with the benefits of running a model that was trained on everyone's creations.
I don't need software developer salaries to go up. That would be kind of selfish and narrow minded.
What I need instead is something that takes the burden off my entire society and gives them a breather. Universal health care to start. They could also use a higher minimum wage, and lower housing costs.
Is it more selfish and narrow minded to wish for a "utopia" that is economically unsound and happens to be your personal preference, or to wish for productive workers' salaries to increase - something with an actual track record of improving any society it occurs in?
All perl programmers should be wishing for ponies, that's definitely less narrow minded.
It doesn't sound like utopia to me, hence the quotation marks. Eminently achievable, but not actually good. Only those engaged in utopian thinking - with a heavy slice of ignorance of basic economics and history - would think it is utopia or leads to it.
Universal healthcare is very sound economically. Costs are lower and outcomes better than under private insurance, and overhead is dramatically reduced.
This is not true, the Kings Fund publishes a report that the Guardian fauns over whenever it comes out because it shows how "cost effective" the NHS is, yet if you read it you find that actual health outcomes are generally worse than other, insurance based systems. Give me wealth and health over a postcode lottery produced by utopianists.
The economically sound thing is to accrue more power to those with wealth. The owners will have access to a machine that turns money into money without a cent flowing outside of the owner class. That'll improve society /s.
>if you find some way to make software developer salaries go up
This is quite easy. Just optimize the models to do reviews and bug finding. This would make developers (who normally hate reviews) quite happy and let them do more coding, thus delivering more value and possibly earning more...
It’s often lamented that some employees have a difficult case to argue for their impact on the bottom line, and as a result probably get paid a lower fraction of their value to the business than other roles where the link is easy to measure.
I can at least “imagine” a model that tries to crack this nut.
But your value to a company doesn’t just come from your impact, but how tough you are to replace, how much others value your skills, etc.
Nike’s logo designer was paid $35. One model says she should’ve gotten hundreds of thousands of dollars, because of what her work product went on to become. Another model of the value says it was worth $35 because that’s what she agreed to.
If, as an employee, you think you’re massively undervalued for the impact you generate, go out to the market and either get another job or start your own business making widgets - either you’ll get that pay bump you expect, or you’ll see you actually were relying on a lot of other supporting mechanisms to generate that value.
The intrinsic satisfaction of increasing the wealth of shareholders. We should all be happy to devote ourselves to getting them more, nothing is more important than that.
Of all the possible criticisms that's the one you chose? If that's the worst of the problems you can see, why don't you buy some stock and became the shareholder. Per your own words, you will get more.
> Of all the possible criticisms that's the one you chose? If that's the worst of the problems you can see
The point is there is little benefit to these technologies to the consumer, especially in relation to likely harm in other areas (you lost your customer service job, but AI overview will answer your trivia question with slightly less effort). Note: little does not mean none.
So the farce is they benefit by religiously worshiping capitalist shareholders.
> why don't you buy some stock and became the shareholder. Per your own words, you will get more.
LOL. Don't you get it? The kind of smallholdings of shares available to regular people won't provide the kind of returns to mitigate any of these harms. They work as a ploy to trick dumb-ass workers into identifying with capitalist tycoons (e.g. opposing pro-worker things that'd get you a dollar more an hour in wages to get a penny more a quarter in dividends, it works because most don't do the math).
> Yes, yes, workers of the world unite. It worked so great the last few times it was tried. You are very smart.
No, I guess I wasn't smart enough to realize there are only two options: the present day status quo or Soviet central planning. Nothing else is possible.
My kids like to use AI to discuss things they learned in school in greater depth, and from different angles than they learned in the textbook. They can also ask "What if" and "Why not" questions from this infinitely patient teacher.
At least with search engines, or even libraries, you're aware that there are many authors of varying reliability and the publications/sites might not be reputable.
AI chat bots will summarize the top N web search results as if they're fact, weaving them into seemingly coherent narratives, all while reassuring the user that their questions are really good and they're learning a lot.
Also it's better not to answer, but flip the question back and let your kid think through it, offer hypothesis, and so on, helping him problem solve, recall, and all that.
I guess you could argue that there should be cheaper software, but most software people interact with is free/ad supported. Where it is paid, it's already a race to to the bottom.
Basically consumers don't really pay for software in the first place, and the leverage from labour companies get through software is already through the roof even before AI. Will much change for consumers of software?
> An expensive AI which simply takes your job or forces you to work harder
But this implies higher productivity, no? This must mean more outputs that should benefit someone, unless the jobs that are being automated had little value to begin with. Seems paradoxical.
There is a practical upper bound on how much labor can be replaced before deflation becomes a problem. AI firms risk spoiling the pot if no other business model is discovered.
At that point it really does sound like ww3 started from the same causes as ww1 - nobody will win, nobody will know why they are fighting, and most of the fighting will be drones being slung over trenches.
reply