foobiekr

The new conditions almost surely being like the old conditions: slavery, sexual exploitation, etc.

He is absolutely a fraud. He has been lying about many things for more than a decade to boost his stock. He has more in common with Trevor Milton than anyone else.

These people are insanely powerful forces in the modern world. Of course we should talk about them (and usually how Wrong, shortsighted, and self-serving they are).

You realized they were always shitheels. Musk was a complete visible fraud long before 2016.

The reason he and Musk are anti-introspection is that when they do it, it hurts. Because they are terrible people.

Better to just not think about it.

I work for a 300+B company that spends nearly $1b a year on AWS.

Microsoft engaged in a relentless romance campaign with our loser EVP and one of his reports for months giving him the cool LinkedIn post opportunities that weak executives crave.

Eventually he started pushing engineering to move to Azure.

We have not yet (many bullets dodged so far) but it’s there and a periodic major time sink entirely due to manipulation and flattery.

The entire “multicloud” push was a marketing effort by Microsoft to try and undermine exec faith in their “what? No, that’s a shit ton of work with zero return on investment” engineering teams.

It's a reinvention of SVG with all the same problems. JSON is not any better in terms of data efficiency or fast updating. It's like these people have never done graphics programming.

BGP attacks have nothing to do with spoofed peers. They have to do with accepted peers behaving maliciously in terms of the AS Paths they advertise.

Once you control BGP you control any IP and can subvert certificate issuance that effectively uses IP to validate certificate issuance requests. For example anything that relies on a file or dns at a specific IP. Once you have done so, you ARE the site, no matter what HSTS says.

We’ve tried to solve this problem a few times with certificate pinning (dangerous) and more recently just giving up and using certificate transparency to try and mitigate the blast radius by hoping the duration can be curtailed. The whole system is incredibly fragile.

As an aside, BGP should move over to TLS (not https, http is a terrible protocol for this) for other reasons (it’s a better option than tcp aom/md5). That this is not already the case should inform people’s opinion of where this stuff is on the security timeline.

You are, of course, ignoring the production costs of the static content that OpenAi is stealing.

Stop justifying their anti-social behavior because it lines your pockets.

99.9% of these would just be malicious spyware that people are tricked into agreeing to.

You may not recall the crazy era of OOP where people would go bonkers with massive object trees trying to objectify everything and using operator overloading to do (dumb) things like adding a control to a window with +=.

More AI bullshit and hype is good for Nvidia. Until it isn't.

I no longer believe this is like the dotcom. Now it feels like the 1983 video game crash.

I have been checking organic and inorganic chemistry skills in ChatGPT pro and it is absolutely, laughably bad. But it sounds good, plausible but it comically wrong in so many ways.

Maybe you should think twice about whether the health issues advice it is giving you is legitimate.

You seem to have missed this part:

>a foreign nation state would pressure a manufacturer in their jurisdiction to include a backdoor

That absolutely is about jurisdiction and is a much bigger, more scalable attack than intercepting and installing implants. More to the point, it can be done at _any time_ not just the initial ship.

It doesn't help that almost all of the big tech companies talking about 5 9s are lying about it; "Does it respond to the API at all, even with errors? It's up!" and so on. If you spend a lot of time analyzing browser traces you see errors and failures constantly from everyone, even huge companies that brag a lot about their prowess. But it's "up" even if a shard is completely down.

The five nines tech people usually are talking about is a fiction; the only place where the measure is really real is in networking, specifically service provider networking, otherwise it's often just various ways of cleverly slicing the data to keep the status screen green. A dead giveaway is a gander at the SLAs and all the ways the SLAs are basically worthless for almost everyone in the space.

See also all of the "1 hour response time" SLAs from open source wrapper companies. Yes, in one hour they will create a case and give you case ID. But that's not how they describe it.

A lot of the AI companies visibly suffer from the engineer's disease. It's kind of interesting to look at them through that lens, and of course, the claims they make about the future.

This is not entirely dissimilar to what Cerebus does with their weights streaming.

Fantasy buildouts of hundreds of billions of dollars for gear that has a 3 year lifetime may be premature.

Put another way, there is no demonstrated first mover advantage in LLM-based AI so far and all of the companies involved are money furnaces.

Windows on a 80486 vs. those boxes felt very much like if you were to compare the latest M5 Macs to, say, a ppc 604 Mac.

No comparison at all. Just every single interactive aspect of them was worse in every possible way and that includes I/O performance. At the time, in that era, people would babble about how much faster SCSI was, but the disks sitting in PCs were blazing fast in practice despite being attached by glorified joystick ports.

That is exactly what the rest of my message said, had you read it.

Some of the problem in California is that the population of want-to-buyers is all focused on a small number of extremely geographically constrained areas.

Their answer is always "more density" but more density isn't going to solve it - the demand is not fixed and density will not alleviate it, so there is natural pushback on ruining the area and not actually solving the problem.

The real answer here is that the valley needs to extend further south, as in the 1990s plans for companies to start building campuses in Silver Creek or further, which would require the government to incentivize such moves.

Had the dotcom lasted another two years or so, IBM, Cisco, and many others would have done the capital investment in campuses further out, and the viable housing area would have increased dramatically, but it didn't work out that way.

Why would you build more cars as prices fall?

What do you think homebuilders do? Just close up shop when the market dips?

The answer is there are companies in the business of building homes. If there is profit to be made, they will keep building. Most of the issues we have today are caused by extreme and dysfunctional regulatory regimes that destroy the return on investment by extending the process out. More time sitting on the property before building = lower investment return, and other investments end up making more sense.

The type of person who funders fall for is a personality type. That they have failed in the past doesn't matter, they fall in love anyway. This time is different.

Venture guys aren't as smart or analytical as their propaganda would lead one to believe. A lot of them are just people who got lucky once.

For now.

the main value is that it's way easier to make an emulator of a console than some point-in-time windows PC.

There is an irony that Wine is the most stable linux ABI for GUI applications in 2026.

I managed a lab of them. I _hated them_. They were unreliable, slow, and just absolutely miserable because they created endless complaints.

We were rolling out labs of Windows machines. Except for the lack of terminal, they were better on every single axis for the common university lab use cases - mostly netscape/mosaic and applications..

I also managed NeXT slabs and cubes; they were vastly better than the sun boxes because we had installed HDDs in the cubes and extra memory. The only problem with them was the absolutely terrible, shit behavior when users accidentally browsed the AFS root...

The only positive thing I can say about those Sun boxes is that _one_ behavior was better than NeXT. With NeXT, students would pull the power on them after wating four or five minutes of the beachball due to AFS I/O.

It's really probably more about wafer and chip allocation.

In 2000 I worked for a company that was building a mobile telephony and data product. The partner company asked us to help them implement the lawful intercept function, as is required by law, which we did, however they were asking for 5+% LI traffic when the common practice was 2-ish%. Our hardware was exceptional, we could trivially have done 100% at line rate with zero impact. The engineers all stepped aside, and finally: "Fuck those guys. They get their 2%."

It's one of the better ethical moments I've had in my career of working for _mostly_ very ethical companies (so obviously not any social media or crypto).

this principle applies to a lot of things. signaling for example. optical links. oldschool optical links (OC48 timeframe) did not feature scramblers and so a malicious packet could on occasion cause them to de-train and go out of sync since it's extended loss of light.

long since fixed but a common problem.