Hacker Newsnew | past | comments | ask | show | jobs | submit | rowanG077's commentslogin

Of course they can allow it. They choose not to. They choose to screw over all users because they are afraid of some company making a claude ripoff. It shows a lack of faith in their own engineering. It shows a lack of respect for users.

I don't think there are public numbers. No doubt IBM knows. I do expect that trend to reverse this year if true.

https://www.lto.org/2025/07/lto-tape-technology-shipments-sc...

https://www.theregister.com/2025/07/23/lto_2024_tape_shipmen...

Unless something is wrong with these numbers, it's simple enough to do rough math to get tape count and compare with historical numbers.

Claims of 5.7 or 5.8 million drives over the lifetime of the format can also be compared to older data to see a slowdown.


I also find it so weird to play this on the person of Altman or Amodei. These are basically fungible public faces. If they die this very moment AI progress wouldn't halt. I don't think it would even be impacted. If anything you should be mad at governments not legislating if you are anti AI.

The ship is going where it is because of the captain. If they die this very moment, the ship will not go back.

And yet,


Using a mid-IR array with sub 10nm resolution is anything but an engineering path. Tech like that has never left the lab afaik.

Fair point. That's why the paper labels it Tier 2 (near-term research) rather than Tier 1 (existing instrumentation). Tier 1 — scanning probe read/write on a single sample — is the immediate validation target and requires no new technology.

That's pretty awesome. Dead locks are extremely tough to debug. There are even cases where I saw behavior in code that might have been a dead lock. I never found out though.

If you can get a stack trace of the process it is usually pretty easy to figure you will see two or more threads waiting on different futexes. I would say async deadlocks are far trickier to debug - task a sends a message to task b and task b sends a message to task a. Both tasks wait on processing further messages until the other side replies to their message. At the end of the day you need strict ordering of who is allowed to call into the other and block.

I don't think describing them as friends is entirely correct. People give money to people they trust. And friends often are in that subset of people. But that's not a strict requirement.

They trust people who look and smell like them or the people they golf or drink with or are part of the same fraternity or tennis club.

I'm not sure what your point is. Of course people who see and observe others on a daily basis in the flesh can determine much better whether they are trustworthy or not. They sure as hell don't think some random person who has no credibility is trustworthy.

The point is the definition of trust is flawed if what you're trying to measure is technical impact and quality or ability to execute?

It's literally in TFA.

Thanks, admittedly, I hadn't read the article at the time.

I was thinking: "No way this has existed for decades". But the earliest I can find it existing is 2008. Strictly speaking not decades but much closer to it than I expected.

Reconnection does not look impossible. But it will be extremely hard.

I think what is much more intractable is actually massive amounts of axons you'd need to reconnect, and you'd need extremely good classification to connect the right axons from host body and brain together. I think the only way to do it is to coax the new body/brain combo into self-repair.


Having had a need for nano positioning in the past I really try to avoid PI like the plague. Bad and slow sales people, opaque pricing, no way to test their product before they expect you to lay down 50k. Bad integration with other tools.

Pure hardware product is great though I admit.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: