AYBABTME

Where does defense stop in an integrated world? Aircraft carriers and a strong navy are necessary to protect waterways and freedom of navigation, which the peaceful side of USA absolutely needs in its pursuit of freedom, peace and liberty. East Asia's history is a great example of what happens when you focus only on your internal world and close off to the rest of the planet. Also why they're not falling for the same mistake again and not having these qualms in the modern era.

Peace requires policing, a credible deterrent to abuse. On the world's stage, there's no actual hegemon for the planet. You gotta choose who will police by mere "I'm stronger than you". I think the US is the most acceptable police, at least it's an elected democracy with individual freedom at its core.

At the end of the day, a government is whoever is strong enough to prevent the other one from governing: provide protection and services to your people in exchange for being the boss. Roaming bands overthrow each other, still visible in Afghanistan and Syria, and dictate who's effectively Government on a given day in a given area. Who applies the laws, who provides justice, who keeps you safe. Scale that up to modern nations and we look at it as if it was the natural order of things: just stick to you backyard and all will be good. But no, it was never like that and even less so these days: the world is shrinking rapidly.

Can't have peace without deterrence against bad actors. And deterrence doesn't stop only on your turf.

Doesn't seem to work? It just spits out a fixed answer about RAM. qwen3.8 27b

All this stuff was in motion while Tesla was in its COVID boom. It's hard to believe that this is all some PR stunt due to Tesla's current market posture, when the Cybertruck, Roadster, were all specced the same in 2021. It's not like they suddenly came up with these gimmicks. The gimmicks were there from day one.

Not sure why the author thinks Anthropic's models spend more power and water than DeepSeek, there's no evidence of that. Their pricing has more to do with premium perception and less to do with COGS.

Everyone does optimization of model serving because it's good for every player in there.

(Also the water consumption thing is not a real issue.)

A thought provoking leading question I have is (1) a space heater throws power into a coil and generates heat and nothing more in the world (2) an LLM into a GPU throws power into the GPU and generates heat and organized information.

Is (2) generating less heat so that the laws of thermodynamics account for the increased amount of "useful stuff" generated (the organized information)? Switch (2) for a bitcoin miner if you prefer.

Is the act of increasing order in the data, somehow costing some energy? Is there a fundamental minimum cost to flipping bits? Something about the switch of energy to matter and back, seems involved here.

I think the nature of computation is holding some secrets about the meaning of the universe, if there is any.

There is no such thing as a general order; if people accept this then the problem relaxes and we stop trying to design impossible things.

Nodes shouldn't be observers. End users should be the observers. The ordering of event only matters from the perspective of the "I'm adding this to my shopping cart and clicking the buy button and then it said I bought it" story being internally consistent. The observer here is the end user and the end user knows what happens-before relationships were.

Databases should flip this ordering inside out and let the observers (or proxies for them) report what they did in which order. And it just happens that protocols for doing this already exist: distributed tracing encodes these relations.

A meaningful global order doesn't exist and attempts to conjure one only conjure a meaningless one, or one that throws away a lot of turns trying to perfectly order everything from some arbitrary internal observer.

In most of my multi-agent workflows, I always end up asking them to create a group-chat system to coordinate and post updates. I thought this was such an obvious day-0 discovery that I assumed it's a well known and understood pattern that wasn't worth talking about. I once again discover that what I take for obvious and granted, might not be.

I've found that similarly to how teams can degrade into spending more time bikeshedding and on the watercooler than on work, agents also tend to end up spending way too much time coordinating as opposed to doing the work. And so I rediscovered that it's better to have one agent that's the Manager (on a Manager Schedule) and the rests be builders (on a Builder's Schedule), where the manager might be interrupt driven, but the builders need to be able to focus for a while without interruption (context poisoning).

Thanks for writing this and demonstrating that writing about anything is useful to share knowledge and practices. In the end, I learned a lot from Martin Fowler and his gang and I guess I should pay back and write about my own discoveries, however trivial they seem to me.

I think they're doing this so they can have months of advanced notice of how the protocol will look, to post-train their models to be the only ones with the ability to interact with this protocol on the day it becomes public.

It's like creating conditions for a race where they structurally have months of advance. And realistically they have a credible shot at establishing the standard due to their weight in the industry, and control of the most popular AI distribution method (Claude).

So they create the race, knowing the race is what will likely establish the standard, and using some pretext to prevent other competitive athletes from training for it in advance. Other athletes only get to start when the gun fires.

I don't get how this is unique to GPU clusters. As a general rule, underwriters are not qualified to operate and maintain the assets they underwrite loans for. That's why houses, cars, equipment, ... go at auction at a fraction of their value. And why lenders have insurance.

The better parallel is "why did Google make Kubernetes open-source" or "why do large for profit entities engaged in competition, use open-source as a strategy against their competitors"?

"Commoditize your complement" -> https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/

I wrote "lightweight polices" not policies. The police presents itself as benign looking in a public context. Enforcement of day to day offences is done mechanically by machines. A state trooper doesn't stop you on a speed check with his hand on his gun.

Yes, online policies are wild and not lightweight at all.

Something missing as cultural context is that deepfake, involuntary "porn", and all sorts of abuse of personal image, are a rampant and omnipresent problem in Korea. Many things are great here, but the sexual landscape when it comes to men versus women and kids, is nasty. You can't really apply a Western mindset to this without understanding just how messed up some of that stuff is. So whatever you think of the mechanism, the problem behind it is very real.

I do think a proposal that AI-filters content on small forums is a bit weird, and probably clumsy. But Korea faces a real problem and usually leans toward a bias to action and "just do it". It leads to weird stuff but also to dynamic problem solving. The part I'm trying to preempt here is measuring this against so called "universal" values; these French Revolution/Enlightenment ideas of universal rights aren't really universal, they're one culture's logic, consistent inside its own bubble but exported like it's the default for everyone. I'll say, I do like them. But other self-consistent logics exist, and I think Korea's set is one of them. It's going to sound cliché but it leans on harmony and the group where the Western one leans on the individual. Both produce aberrations, only different ones.

For example, first time I came here I thought it's crazy to have so many speeding cameras and CCTVs everywhere. Years later I didn't so much "got used to it" but I think it's a tradeoff that mostly works and I grew to appreciate it.

Korea prefers lightweight polices (literally friendly looking) with a lot of automated, bulk enforcement, instead of sparse enforcement backed by the occasional armored truck. That's a design choice, not a slide into dystopia.

So all I'm trying to convey is, keep an open mind, and don't apply some supposed "universal" mindset blindly. Critique the mechanism all you want. Just don't do it by treating one culture's values as the yardstick everyone else gets measured by.

Fwiw I think it's a misfire. But I don't think it's a slippery-slide down dystopia. It's just Tuesday.

TLA+, P, Lean... formal methods and previously esoteric testing methods (property based, mutation... testing) should become the default. I think it's the only way we can really reap the benefits of agentic coding.

I wrote about this a bit on my blog[1], different angle but along the same line. You explain TLA+ and model checking well which makes the case concrete.

I'm curious of you have thoughts on these other methods and tools like P, Lean, Dafny, etc?

[1]: https://aybabt.me/blog/correctness-agentic-world

That's a lot of very large jumps to come to the conclusion that Bun isn't going to turn out well.

The problem IMO is that they filled GitHub with Microsoft folks who just don't have the engineering self-sufficient hacker culture that is required to balance the "attraction park" vibe that GitHub paired it with. So now it's just an attraction park for Microsoft employees to go and do silly work with teams of 100 that should have been done by a skilled team of 5 hackers.

I was there for a couple years after the acquisition and just couldn't stand seeing it. I felt I was becoming useless working in a mad house that was becoming more maddening everyday. And MSFT just keeps replacing leadership with more and more disconnected people who just don't get it, who just never used GitHub like the OG users did. Two years ago I interviewed again for my old team, largely out of curiosity, and the Microsoft engineering manager asked me some brain teaser question as my interview. The disconnect is just too large.

They don't take GitHub seriously. It's a toy to MSFT and vibes matter more than the product itself. And they hire for it using MSFT drone logic, fill it with people hired and profiled to be MSFT-lifers, and these two things don't mix.

Sorry I don't have anything great to say. And of course, many of these MSFT folks were actually damn good, but they were swimming in a sea of MSFT drone.

This reduces writing to one concept: thinking and the writing is just a byproduct. But writing is also presentation and also communication.

There is nothing wrong with speechwriters. Various authors spilled out their thoughts in rough format and had writers turn them into better structured, prosed and understandable projections. Hand writing each sentence that is presented as an end-product to the reader doesn't solve that problem.

Forcibly coupling the two is an arbitrary choice that may be a valid tradeoff for some and not so for others, and not so for _all_ writing.

I'm not good at looping through a document with proper english prose. My writing is raw, particular, and I gloss over a lot of details. LLMs help me turn my shitty extensive notes in bad grammar and syntax, into shareable and understandable artifacts. They help me turn more of my thoughts into ingestable communication by others. Without AI, I communicate less of my thoughts due to friction. My thoughts are formed and authored and written, but not in a format consumable by anyone else.

Ebikes help older riders keep riding.

Everything you say is probably true and I agree... and yet.

What matters is not just what you plan to do, but what your opponent thinks you'll do. The US in general believes that China wants to invade or control Taiwan in some way. This mere belief is sufficient to cause it to take action.

I mostly agree with everything you say, I just see the balance lying elsewhere on the spectrum. I think China is on it's way to securing its energy supplies with renewables but not quite there, and that the US is taking this window of opportunities to do what it can to attempt to degrade China.

Whether China plans to actually invade or blockade Taiwan or not doesn't matter if the US thinks it will. AFAICT the US is convinced it will, and the mere threat of this is enough to justify Venezuela and Iran, I believe. Higher oil prices are less worse than no more semiconductors.

And I think Russia might have gained some territory, but at the cost of being completely sucked into the conflict, having lost strategically by (1) being unable to support and defend its proxies and (2) having its arsenal and technology thoroughly analyzed and proven ineffective against US weapons. All actors involved know this and it will not remain, but until solved this means that the US knows it can strike countries defended by Russian weapons, at least until counter measures are researched, developed and distributed. This is a temporary advantage and moment of clarity that lasts a few years, not a sustained advantage.

The risk of the US being equally sucked into Iran and suffering the same fate is very high. And China's best strategy here is probably to sit and wait and help US opponents keep the US busy for a while, like the US did on Ukraine with Russia.

(1) China is more sensitive to shipping and oil shipments (and derivatives) than the US. It hurts both but China much more. The US is hurt in so far as high gas price are bad for elections, a small price to pay for strategic advantage.

(2) Iran has a temporary but unsustainable interest in constricting traffic, and it's not the only country who could impose a filter there. The mere credible threat of a strike on shipping is enough to stop it, so other countries basically have an equivalent capability to restrict traffic. And all countries, including Iran, are unable to sustain a prolonged closure. The current situation is an unstable, non-equilibrium situation for Iran and it's neighbors.

Overall, all of it doesn't really matter to the US because simply taking Iran off the supply chain of China is good for them. They spin the narrative about starting the war for a variety of other reasons so that they can justify the pain it inflicts on their allies (Korea, Japan - very dependent on those hydrocarbons too, and EU) and choke China's oil supplies without looking intentional. Last time the US overtly blockaded an asian nation's oil supplies, Pearl Harbour happened.

Which is another reason why China had such a structural incentive to move toward solar power, battery storage and renewables in general while also powering most of their early growth with dirty power plants.

I think Trump wants to be remembered for having neutered the China threat and having restored American supremacy and dynamism, and doesn't care too much about what it will cost at the next federal elections. I think he cares more about his legacy and wanting to be remembered as a historical figure on the strategic level. He's portrayed as being merely a fool with self interested dictatorial tendencies but I think attributing such simple intentions to him is self deception and leads to poor analysis. It doesn't pay to trivialize figures for disliking them or their actions.

Without taking camp here, I'll say that taking Trump for a fool is shortsighted, in my opinion.

I think this analysis is missing the big picture.

(1) reducing oil shipments to China is good posturing for the US; hence Venezuela and Iran ahead of 2028. These are shaping operations. China suffers more from these conflicts than the US.

(2) Iran isn't the only one who can control passage throught the strait. All gulf countries can do so. If Iran can cheaply cut off passage, so can Saudi Arabia and UAE and everyone else there. They all have a long term mutual need in keeping this strait open.

All these recent analysis of conflicts in isolation, which always assume a lot of self-interest in disliked politicians, seem to make the analysts and authors blind to a much more probable and sensible grand strategy. Russia invading Ukraine and failing, has been the greatest strategic gift Russia could give to the US against China in setting the stage for shaping a defence of, and deterring an offensive on, Taiwan. Russia lost the ability to defend its proxies at a cost asymetrically small to the US. Hamas broke rank and allowed Israel to eventually decapitate Iran's proxies and air-defense step by step instead of all at once, setting the stage for the opportunity of the current war. And Russia being distracted also gave the US carte-blanche in Venezuela, not only via distraction but by proving that Russian air-defense isn't the thread it was thought to be.

The remaining strategic tension, in my humble opinion, is whether the US depletes its stockpiles too much without a caught up manufacturing capability, so that a Taiwan conflict becomes easy to win by default for China (via a blockade which would essentially be a cold war with few deaths and minimal damage) or if the weakened China (due to oil constraints) would be simply unable to attack in 2028, the strategic window when it can do so.

The situation, in my eyes, is evolving in a state where only two modes become dominant and both are slightly better for Taiwan.

This right now today is making the case for OSS AI and local inference. 200$/m to get rate limited makes a RTX 6000 Pro look cheap.

Let's say 100k files is 300k syscalls, at ~1-2us per syscall. That's 300ms of syscalls. Then assume 10kb per file, that's 1GB of file, easily done in a fraction of a second when the cache is warm (it'll be from scanning the dir). That's like 600ms used up and plenty left to just parse and analyze 100k things in 2s.

If a human asked me this question, I would be confused by the question as ambiguous since it suggests something odd is implied but underspecified. I think any confident answer either way by AI is lacking in pedantry.

For some reason, I always found the arguments for "it's better to not know" for these tests to be strange and slightly infantilizing. But of course this must not be the end of it, and there might be some more well thought out arguments from bioethicists that go beyond "the patient can't handle the truth". Because this argument seems like it's doing a lot of heavy lifting without much evidence.

I think it's a mistake to believe that this money would exist if it was to be spent on these things. The existence of money is largely derived from society scale intention, excitement or urgency. These hospitals, machine shops, etc, could not manifest the same amount of money unless packaged as an exciting society scale project by a charismatic and credible character. But AI, as an aggregate, has this pull and there are a few clear investment channels in which to pour this money. The money didn't need to exist yesterday, it can be created by pulling a loan from (ultimately) the Fed.

the rebuke is that lack of chaos makes people feel more orderly and as if things are going better, but it doesn't increase your luck surface area, it just maximizes cozy vibes and self interested comfort.

It feels like we're doing another lift to a higher level of abstraction. Whereas we had "automatic programming" and "high level programming languages" free us from assembly, where higher level abstractions could be represented without the author having to know or care about the assembly (and it took decades for the switch to happen), we now once again get pulled up another layer.

We're in the midst of another abstraction level becoming the working layer - and that's not a small layer jump but a jump to a completely different plane. And I think once again, we'll benefit from getting tools that help us specify the high level concepts we intend, and ways to enforce that the generated code is correct - not necessarily fast or efficient but at least correct - same as compilers do. And this lift is happening on a much more accelerated timeline.

The problem of ensuring correctness of the generated code across all the layers we're now skipping is going to be the crux of how we manage to leverage LLM/agentic coding.

Maybe Cursor is TurboPascal.

Hydrocarbons can be synthesized.

edit: let me elaborate.

My point is that the chemical complexity (manufacturing uses) can be reproduced, and the energy storage density also can be. So really the gift of hydrocarbons under the ground is more that readily available energy is under our feet to help propel us towards higher levels sources of energy. IMO it’s a stepping stone and that’s effectively how humanity is using it.

You can chose to see it as astroturfing, or see it as people actually thinking the superlatives are appropriate.

To be honest, it makes no difference in my life if you believe or not what I'm saying. And from my perspective, it's just a bit astounding to read people's takes that are authoritatively claiming that LLMs are not useful for software development. It's like telling me over the phone that restaurant X doesn't have a pasta dish, while I'm sitting at restaurant X eating a pasta dish. It's just weird, but I understand that maybe you haven't gone to the resto in a while, or didn't see the menu item, or maybe you just have something against this restaurant for some weird reason.

The fact that as many engineers are on payroll doesn't mean that "cloud" is not an efficiency improvement. When things are easier and cheaper, people don't do less or buy less. They do more and buy more until they fill their capacity. The end result is the same number (or more) of engineers, but they deal with a higher level of abstraction and achieve more with the same headcount.