I guess this is it. I wasn't able to force myself to listen to the whole thing, but then I'm an old fogey.

https://www.youtube.com/watch?...

So the 'hyperloop' people have a cool website; while the 'train' people are just plain getting on with building stuff; whether conventional or the now-quarter-century-ish old maglev option.

Looks like someone signed up for another round of 'faff with apps vs. offshoring our entire high tech supply chain' and hoped it would work better this time.

And some dumbass 'managing director' is telling us that a gigantic safety-critical vacuum system is 'not effected by strikes'; more or less because he has no idea what the maintenance and operations would involve? Truly a joke telling itself.

It's easy to have unique keys in your spreadsheet so that you can easily relate information on different sheets to one another. The problem is, actually doing the processing that a SQL server would do trivially is irritating, and then it will be processed slowly every time. Whatever Excel does or doesn't cache, it isn't enough. You can do big complicated things, but they work slowly, and maintaining it is irritating at best. When you do complicated things either your formulas get long, or you wind up having to write code, or in fact often it's both. At that point you're way better off IMO doing it in something else so that at least performance is good when you're done, and you never have to screw with editing a long formula.

That's exactly why it should be in a database and not a spreadsheet. Spreadsheets are best when you have a reasonably limited number of columns. It's also a horrible PITA to use them as a relational database (it's more or less possible, but you don't want to do it) so hiding pieces of that complexity in other sheets in order to limit the data the user interfaces with on the main sheet is just a lot of extra work you wouldn't have to do if you used another solution.

It doesn't really matter where it fails, if Excel can do it and Sheets can't then Google has to admit inferiority to Microsoft which is never a good look.

Biden went around congress to fund a different genocide. Pretty words, but living up to them is another matter.

The UK is one of the halfway countries that still uses miles and gallons for vehicles

Only for fuel consumption, but not fuel. Which his funny because we dispense fuel (and just about everything else) in liters, so MPG is an entertainingly irritating figure. We should use MPL obviously.

Pandering to Reform voters is pointless, they are not going to vote Labour unless they go full fascist.

Probably not even then. If you're going to vote for a racist, why not go for the real deal instead of a half arsed knockoff where you're not even sure if they're really committed or just doing it to try and impress.

Here it is every single year. Varies by State, of course. Some States it is 2 years, some have no inspections at all. https://en.wikipedia.org/wiki/...

In NM it's yearly: they ensure your tire treads aren't too deep and that you have one emergency spare in place of a regular wheel. They'll also put some chips in your windscreen if you ask nicely.

it augers the beginning of the end of total immunity for social media conglomerates.

You've drilled down to heart of the matter.

They do. Some people don't use them; and (if disciplined) use one or more worksheets to store data and refer to it purely internally and (otherwise) just sort of ad-hoc mix data and formulas.

In some cases a database connection is where the data comes from; but the number of cells grows because it's conceptually easier(and in practice often less opaque, given the ugliness of displaying very large cell contents) to munge on the data step by step rather than trying to ram everything into one transformation.

Coming from the IT side; and having to field questions from the perpetrators of some absolutely hideous excel sheets from time to time(no, I didn't even know that there was a way of creating a type of embedded image that actually quietly triggers the print spooler subsystem to do something that generates a new image based on the contents of another region of the spreadsheet, still don't know how they did that; but it's objectively depraved) I understand the hate; but I do have to admit that spreadsheets are pretty good for napkin-math thinking-it-through type processes.

Like when you work it out on paper; you've got your input, then you have a cell with the contents of the first transformation you wanted to make, then the second, then the third, and so on, and at each step you can think "does this value make sense?"

It rapidly gets out of hand in quantity; but as a rapid sketchpad for thinking something through you could do a whole lot worse. It's also tempting(again, tempting down the path of darkness in quantity) for dealing with jobs that need both a bit of string munging and a pretty-printed output.

You send the intern down to storage with a barcode scanner and have them start snagging SNs and MACs and stuff from the shipment of new gear. Turns out various vendors use different prefixes on different barcode values to inform their own ERP/inventory system/warehouse people which of the 5 closely spaced barcodes their scanner hit. And each vendor uses a different set of conventions, and while obvious enough they aren't documented. Ok, no problem; intern comes back with raw list; all the Lenovo SNs get a 'last x characters' substring; all the Cisco MACs get another transform, whatever.

Obviously if it were your inventory/warehouse system you wouldn't be treating the barcode scanner as a raw HID device and doing ad-hoc transformations, there would be a program that automatically uses the prefixes to populate the correct parts of the form; but you want to stick your head into ERP project hell rather than come up with maybe a dozen lightweight string manipulations? Obviously, you could also do it in your choice of scripting language and iterate through one CSV to create another; but that mostly just conceals what you did from anyone who doesn't use that scripting language; while you can walk basically anyone employable through the logic of the spreadsheet prettifying.

or a paper in Nature

That's not a paper in Nature, it's a paper in Scientific Reports, a much lower tier journal published by the same company.

Hells bells that is godawful though, and inexcusable from a publishing point of view.

SciRep is a very mixed bag. It's where you go when you've exhausted the higher impact factor journals. There are some really good papers in there, but also bad ones and with the vast deluge (even pre AI slop) of papers, getting good peer reviewers for mid to lower tier journals has been more or less impossible for years.

I read that as H2O conflicts first... I will let someone else insert the other obvious meme.

Yes it will keep being LLMs are LLMs week until the outside world moves on from hammering "LLMs are magic" week every week.

LLMs are of course LLMs.

The problem is that the LLM only does one trick. When you start integrating other software with it, the other software's input has to be fed in the same way as your other tokens. As the last paragraph of TFS says, "every clock check consumes space in the model's context window" and that's because it's just more data being fed in. But the model doesn't actually ever know what time it is, even for a second; the current time is just mixed into the stew and cooked with everything else. It doesn't have a concept of the current time because it doesn't have a concept of anything.

You could have a traditional system interpreting the time, and checking the LLM's output to determine whether what it said made sense. But now that system has to be complicated enough to determine that, and since the LLM is capable of so much complexity of output it can never really be reliable either. You can check the LLM with another LLM, and that's better than not checking its output at all, but the output checking is subject to the same kinds of failures as the initial processing.

So yeah, we can do that, but it won't eliminate the [class of] problem.