I read it yesterday, and the biggest miss is that it makes an unfounded assertion that the human processes and loops can change that quickly. Secondarily, it makes the assertion that the ecosystem can be secured in one year, or that people will just stop caring that they could lose all their money at the drop of a hat. (And somehow crypto will be able to scale in a way it couldn't over the last decade.)

We are still in a ten year process of continually exposing bottlenecks with a large group of bad actors ready to take advantage of any gaps. All it takes is for a deca-millionaire to lose a significant chunk of their money due to an AI mistake for everyone to panic and set everything back.

I'll admit - I don't understand how companies haven't gone cat and mouse with this. For example, if a company wants to create a speculative product, couldn't they just fund a "contracting company" that hires people? If the bet is successful, the parent company buys out the "contracting company" - otherwise, they stop paying the other company and it just goes bankrupt.

I think it's basically "tactical insolvency."

This isn't about turbotax. We are not going to have 100 million people create a tool to do their taxes. Heck, I have the knowledge (and my previous taxes as base) to do my own taxes, but I don't. Why? Because I can't e-file on my own, and I can't be bothered to print something out and wait forever for a refund. (RIP IRS Direct File)

It might be about Quickbooks. However, I think it's more likely to see people build QBO alternatives and the market flooded with vertical-specific competitors than everyone making their own. As Nate Jones says, The cost of software might be going to zero, but the cost of attention is high as ever, and trying to create custom accounting software for a small business is not a good use of attention unless you're already an accountant. (and maybe, not even then.)

I don't even think it makes sense for large companies - even if you can replace SAP or Workday with your own custom software and have it be cheaper to maintain the four developers that do everything for you, it's a bad use of your corporate attention in all but unique cases. Otherwise, everyone would already be using Odoo.

That's fine. And Tailwind never gets updated again, and you have to go find another solution and other companies have to migrate off.

Tailwind generates an enormous value and cannot pay their employees to maintain that value. That's a problem. It's not like tailwind is 200 employees, either.

We need to have a better model for OSS that has product companies shouldering their fair share.

I think they're trying to say that the GPUs will depreciate more quickly than expected and thus the expectations of their return on investment (on which the loans financing the GPU purchases) depend on will leave all of these major companies heavily in debt without revenue generation on the assets to justify their purchase in later years.

Conceivably, this could lead to bankruptcies and a chain of failures from companies like google and amazon, with a massive drop in stock value and a "too big to fail" problem for the US government again.

(This is me trying to explain the argument, not saying I agree or disagree.)

Not if there are too many people trying to do this at the same time.

We'd quickly overwhelm the neighbors' capabilities.

It depends.

If someone is in active recovery, they may exactly know what they can handle at that point. For example, someone who had sexual abuse in their history knows they don't need to see someone else's depiction. That's not going to help them face and overcome it and will more likely set them back for the day. If they're going into an R rated movie, the trigger warning is right there in the rating system. That's not the case for all media.

Or, it might be that on some days they can handle it, and some days they can't. I don't always want to hear about religious trauma because I already have that T-shirt, but on other days and in other contexts I might be curious.

I appreciate trigger warnings though I don't need them anymore.

Unfortunately, my brain has a fairly strict rate limiter and I have yet to see a way to pay out of that.

I would like to point out that Linus is against forking the kernel, and his group essentially demands a unified kernel and toolchain (with different distros having different configurations of these pieces).

[Citation Needed]

Torvalds's copy isn't deployed by most people. Red Hat does its own fork (or patchset), as does Ubuntu. TiVo certainly keeps its own copy. Andrew Morton has gone on record saying that a competing fork would be impractical, but I haven't seen anyone "against" such a thing.

If someone really wants to create a dependent sound system, I'm sure Mark Shuttleworth would like to hear from you if you can make the experience better.

Frankly, for most people, they can just use Ubuntu and forget about every other distribution on the desktop.