Hacker News: adgjlsfhk1

New comment by adgjlsfhk1 in "The experience of rendering Arabic typography and its technical debt"

adgjlsfhk1 — Sun, 14 Jun 2026 01:53:57 +0000

the other part is that numbers and symbols were very much not the priority. The printing press was for books, magazines etc. math remained hand written until the computer

New comment by adgjlsfhk1 in "Statement on US government directive to suspend access to Fable 5 and Mythos 5"

adgjlsfhk1 — Sat, 13 Jun 2026 20:59:17 +0000

The us got those agreements in the first place in exchange for 0% tarrifs for international trade. Now that the US has dropped free trade, there's very little incentive for other countries to hold to their side of the deal. If europe wanted to, they could really go to 25 year on US copyright with a plan to go to 0 if there was retaliation

New comment by adgjlsfhk1 in "Statement on US government directive to suspend access to Fable 5 and Mythos 5"

adgjlsfhk1 — Sat, 13 Jun 2026 13:32:54 +0000

Cory Doctorow gave a talk a couple months ago with the answer to how. Stop honoring US copyright.

New comment by adgjlsfhk1 in "US Government directive to suspend access to Fable 5 and Mythos 5"

adgjlsfhk1 — Sat, 13 Jun 2026 02:17:26 +0000

it may be really good pr, but it's really bad for their IPO. If their market for future models is usa only, their potential revenue is cut by 50% at least. (and it's even worse because it means Europe, India, and China will all have companies making their own models that anthropic needs to stay ahead of)

New comment by adgjlsfhk1 in "Claude Fable 5: mid-tier results on coding tasks"

adgjlsfhk1 — Fri, 12 Jun 2026 05:35:38 +0000

they're worried about people creating bioweapons

New comment by adgjlsfhk1 in "Waymo Premier"

adgjlsfhk1 — Fri, 12 Jun 2026 00:12:52 +0000

The good rule that most of Europe uses is that motorcycles can "lane filter" (i.e. go in between lanes but only for cars that are basically stopped and only at low speed). Going between lanes is suicidal at high speed, but if cars are <5mph and the motorcycle is ~10mph it's actually safer for the motorcycle because it removes the chance of them being rear-ended. (It also makes motorcycles faster than cars which is helpful for discouraging cars in cities)

New comment by adgjlsfhk1 in "We Think the SpaceX IPO Is Overvalued"

adgjlsfhk1 — Thu, 11 Jun 2026 23:59:09 +0000

"If you give me 5 years and everything goes perfect, I can give you a system that's 9% worse than what you already have", you don't have a great investor pitch.

New comment by adgjlsfhk1 in "Can LLMs Beat Classical Hyperparameter Optimization Algorithms?"

adgjlsfhk1 — Tue, 09 Jun 2026 21:26:06 +0000

This is really interesting, but IMO their metric isn't great. By using qbits*gates, they are only able to find interesting points along a specific line of the pareto frontier, but it would be more interesting to look for improvements across the entire frontier (low qbit is especially interesting)

New comment by adgjlsfhk1 in "We Think the SpaceX IPO Is Overvalued"

adgjlsfhk1 — Tue, 09 Jun 2026 04:12:38 +0000

The comparison isn't half a billion for a datacenter, it's half a billion for a big shed (and some solar panels). The problem of space based data centers is that the only effort they save you is building a big shed to put the GPUs in (and potentially cheaper power at the cost of more expensive cooling)

New comment by adgjlsfhk1 in "xAI is looking more like a datacentre REIT than a frontier lab"

adgjlsfhk1 — Tue, 09 Jun 2026 03:26:45 +0000

Not necessarily. The GPU leases Spacex has made are month to month, so they are taking on all of the risk. If demand goes down, they're the ones stuck with the assets.

New comment by adgjlsfhk1 in "We Think the SpaceX IPO Is Overvalued"

adgjlsfhk1 — Tue, 09 Jun 2026 02:38:38 +0000

Part of the problem with this argument is that for them to be worth a lot, you need both the amount of stuff launched into space to go up massively and for no one else to be able to build a rocket that's anywhere near as good. Spacex is doing really well now, but if Rocket labs, China, Blue Origin, or anyone else manages to make a rocket that's half as good, spacex needs to maintain a monopoly despite not having anything especially magic.

New comment by adgjlsfhk1 in "We Think the SpaceX IPO Is Overvalued"

adgjlsfhk1 — Tue, 09 Jun 2026 02:32:23 +0000

Because 50 rocket launches seems hard to make cheaper than a building.

New comment by adgjlsfhk1 in "Apple Core AI Framework"

adgjlsfhk1 — Tue, 09 Jun 2026 02:17:50 +0000

yes and no. We've reached the point where larger models are higher quality, but they're also too expensive and slow to be used broadly. The giant models, however are still useful for training smaller models that are actually deployable.

New comment by adgjlsfhk1 in "New U.S. college grads now have higher unemployment than the average worker"

adgjlsfhk1 — Sun, 07 Jun 2026 00:15:34 +0000

The biggest problem that increasing housing supply can solve is the gap between minimum and median apartment price. In places with very constrained housing markets the cheapest slumlord apartments are very expensive (~70-80%) compared to the price of well maintained apartments. Increasing housing supply doesn't do much to the median housing price (since new houses are expensive), but it lets the price of shitty apartments drop a ton.

New comment by adgjlsfhk1 in "Google will pay SpaceX $920M per month for compute"

adgjlsfhk1 — Sat, 06 Jun 2026 21:44:50 +0000

Even permitting isn't a clear win. You are changing from land permitting (where you can pick the location to be wherever you want) to launch permitting (where you have to coordinate with the federal government for airspace and water closures). Not to mention that with the current regulatory status, a rocket explosion can easily lead to a multi-month mandatory safety review that blocks all new launches.

New comment by adgjlsfhk1 in "Moving beyond fork() + exec()"

adgjlsfhk1 — Sat, 06 Jun 2026 19:48:44 +0000

One os level thing that is interesting to me is if it would be possible/wise to make an OS based on (concurrent) garbage collection.

New comment by adgjlsfhk1 in "Google will pay SpaceX $920M per month for compute"

adgjlsfhk1 — Sat, 06 Jun 2026 19:40:34 +0000

anthropic isn't selling capacity. they're using all of theirs and more

New comment by adgjlsfhk1 in "Google will pay SpaceX $920M per month for compute"

adgjlsfhk1 — Sat, 06 Jun 2026 19:37:35 +0000

that's the best part they might not, and don't really care. they've already made $30b when is the dc never shows up

New comment by adgjlsfhk1 in "Google to pay SpaceX $920M a month for compute capacity at xAI data centers"

adgjlsfhk1 — Sat, 06 Jun 2026 19:34:49 +0000

the math doesn't work. a starlink satellite has ~10kw power consumption. A single ai optimized server rack (GB300) is 140kw. Starlink works because you get a massive benefit from putting networking in space for rural users. no one has made a convincing case as to why putting a data center in space is a benefit that can come anywhere near the drawbacks (inability to service, launch cost, cooling etc)

New comment by adgjlsfhk1 in "Google to pay SpaceX $920M a month for compute capacity at xAI data centers"

adgjlsfhk1 — Sat, 06 Jun 2026 19:07:58 +0000

given that SpaceX is choosing what price they're charging starlink, there's a reasonable argument that starlink isn't profitable either