When Lucas sold Star Wars for $4B, it sounded like a ridiculously low number. Even without the rights to make new content, that was worth well over $1B. If he were shopping around instead of giving it to Disney, it probably was worth closer to $10B. With honest accounting, that's probably in the range of value Disney has received, and that's with having produced the worst possible sequels.

Dell has made lots of acquisitions over the years, and each one had their own systems. Dell acquired me at one point (I'm sure that was their intent, and they just got EMC thrown in with the deal), so I've certainly seen this. It took a long time to really merge the companies, and I'm sure they still have many separate systems. Eventually it makes sense to move everything to the same enterprise management system; at least everything that they aren't keeping separate for an easy future spinoff.

While the obvious answer is to move everything to Dell's existing system, or possibly EMC's, they may well have realized that none of the existing systems were adequate, so it was time for a new one. Might as well rip the bandaid off quickly all at once and be done with it.

Catalyst was the project name.

Or was Catalyst the replacement for the one that failed?

As I recall, the failed one customized everything for how EMC did things. The replacement changed how EMC did things to match the default, as it was easier to change the company than customizes the software. The only pain point I remember on the engineering side was all the part numbers changing to match the new standardized format.

Most of the repairs I've needed on my 2015 Tesla Model S are done at home. Things like replacing faulty door handles (they finally have a design that doesn't fail) and other minor but weird things are done at home--Tesla drives out to me. Same with window replacement.

The only real roadside repairs we've had have been for flat tires or a broken axle (massive pothole). Those usually involve a tow, though we did have Tesla stop by and install a spare replacement wheel and tire for a flat once, which was nice.

We've never run out of charge while driving.

There's just much less that can go wrong that would require immediate service. My car recently warned that the 12v battery was starting to fail, so I got it replaced before it caused a problem, so even that is no longer a sudden critical problem.

I'm thinking combining the retail space with a real distribution center where packages are already sent out from. Or maybe have it near enough with dedicated trucks going back and forth, though that would be less efficient. I'm assuming it's effectively a separate building apart from the merchandise moving, and certainly the retail parking would be separate from the delivery truck parking.

I think it would be interesting to try a physical store combined with a distribution warehouse. The store would have terminals (or you could use your app) where you could look up any item and see where it is. If it's not on a shelf, but in the attached distribution warehouse, you could order it and it would show up in 15 minutes (without extra packaging).

I'm imagining a tunnel between buildings with a long conveyer belt, and items placed in plastic bins with bar codes that would identify the customer, but there are plenty of variations.

Yes, AI will struggle with doing full tasks unsupervised. But it can still do most of the work for many tasks. It just needs supervision by someone who understands the task. Sometimes the problem is the AI making incorrect assumptions about the task (it wasn't fully framed), sometimes as stated in the summary, the AI context window is too small, so it forgets things, and sometimes it just chooses a really bad approach.

I have been using Claude Code a lot recently. It's really good at summarizing existing code. It's good at specific targeted changes. It's pretty bad at designing solutions. I find that while it's usually still faster than doing it manually, I often have to point out where there's a better (usually simpler) solution.

So AI doesn't replace the human, but when used correctly, it makes the human more productive. If instead of having a human do the task manually and compare that to the time taken for a human to supervise AI doing the task, you'll probably find for many that the human can do a lot more with AI. (Yes, I know some studies have shown the opposite, but I think that's mostly people not understanding how to effectively manage AI, which may take some experience and training.)

But AI is far better at almost everything that it was a year ago. So even if it's 2.5% now, it may be 25% next year and 90% a year later. We're living in interesting times.

How many people actually use the speakers built into their TVs? The general rule is you want to get a sound bar at a minimum or a real audio system.

According to Google, about two-thirds are connected to sound systems, which may be right for higher-end TVs. It's probably logical to put better speakers on smaller and cheaper TVs that are less likely to use external sound, but that's the opposite of what manufacturers are likely to do.

We're probably stuck with lousy TV sound unless we get magic speakers that are good, cheap, and small. Until then, we'll have bad, cheap, and small.

You are mad.

My eyes are laid out horizontally, so video should be, too.

If I'm watching on my phone, I'm turning it sideways.

Yeah, I was just playing with my Atari 800XL. I used Claude to write some code for it. So I guess it's still being a problem.

How many doctors will be using this or something like it to diagnose their patients?

If it's anything like my experience with AI coding, it will be great for telling doctors how smart their diagnoses are, but will sometimes point out things that they missed.

Yup. And in many places there are government programs that will subsidize home energy inspections and recommended insulation improvements.

But if you try to improve the building codes to increase energy efficiency for new construction, builders will be up in arms, and it's a tough battle. And that's where you make a huge difference, fixing hundreds or thousands of new buildings at once. And they could also require homes to pass energy inspections to be sold (here we already have septic and smoke detector requirements), which would help a lot for easy fixes like attic insulation.

There's far too much personalizing societal problems instead of working to correct the underlying problems. Both is fine, but the priority should usually be on fixing the system.

A typical 55" TV draws perhaps 80W. Convert to kWh/year, and it's 80*24*365.25/1000 = 701.28.
If you electricity is $0.20/kWh, then you're looking at $140/year for your "art" display. That's over $10/month.

Some people don't care, but it's really wasteful.

Now an e-ink display would be really efficient, only needing power when the image is changed, but those aren't practical for TVs (or likely for any large use).

We have a consistent problem of telling individuals to do better, when that's mostly just a distraction from the majority if the problem. We tell people to turn down thermostats instead of improving energy efficiency codes. Now we're saying blame business class seating when private jets are the biggest inefficiency in the airline industry.

Yes, we can and should look at the small things, but we should really focus our energies on the major culprits.

That's mostly true. I've run some LLMs locally: they're slow and not very powerful, but there are times when having it local is nice for privacy.

I do expect this to change over time, with more operations being practical to run locally. I would really like to have voice input be a local task, for example.