The question is... can you build a body without a brain? I'll bet you actually can't, but our victorian way of thinking about the body doesn't yet know that. What I'm getting at is that our bodies are far more interconnected than we really give them credit. Things that happen in our guts, for example, can have a profound effect on our brains - and quite likely the opposite too. I'll bet that from your head to your toes, your brain is involved in the development, maintenance and operation of all those body parts.

As you point out, our bodies aren't static either - they need to move and work in order to grow or maintain themselves. It's unclear, for example, if you could artificially grow bone that has the requisite density in all the right places to actually be usable. Our bones are subjected to vibrations from all of our movements, and that is the mechanism through which they develop. I'll bet you can't synthesise that sufficiently well (or at least, not easily). Likewise muscle growth - we'd all like to think you could just stick a few needles in and electric-shock them all the grow nicely, but attempts to do that so far have been rather lack lustre to say the least. After that, there are a whole load of things I can't even name which will need doing too. One wonders if it's ever likely to be possible?

FWIW, I suspect more likely would be to rapidly create some of the required tissue for a repair. If you get liver cancer, then they go off and make a load of your liver cells, presumably wash out the cancer and then graft them in place of the old cancerous liver cells. Once they get that working, they might be able to make an entire liver - maybe. But I suspect by the time they want to get that far, they'll either be growing the replacement liver inside another living thing, or else will have worked out if it is ever possible to build body parts without a brain being present.

I saw this come by today... I can't vouch for its absolute correctness as I haven't checked the data they claim to have scraped, but it's a pretty damning indictment on MS's stewardship of Github. From near 100% uptime to quite a lot less than that - I mean, probably not bad enough that you can't rely on it, but you definitely don't want to be in a "I need it back up right now or else I'm screwed" type situation (so don't engineer yourself into such a position).

https://damrnelson.github.io/g...

Microsoft Windows (just Windows, not all the other stuff) is a massive bloat-fest. It's a whole load of interconnected mess, so it's likely impossible to fully test one bit without also testing a load of other stuff too. Then Microsoft also has to test their stuff on a whole room full of different hardware, just to cover their partners, never mind all the other vendors that matter.

There's no way Microsoft can test everything, all the time. It's likely just too big a problem for anything other than maybe a big launch or service pack or something. As such, they likely have to cut down from "fully test everything" to "quick test everything", to "quick test the main bits and call it good".

If there'a any similarity between MS and Oracle here, it's that they've spent decades lumping ever more features into places they don't belong. If (for example), MS had taken a far more modular approach to Windows, then right now, a whole raft of people would be unaffected by this bug because they didn't adopt some lump of services that contains the issue. What's more, MS would be able to internally segregate the problem and so likely would be able to get closer to "test everything, every time" than they can currently, so the likelihood of the bug even getting out would be reduced.

So yeah, there's a wider issue: Far too much legacy software is just too bloated and poorly architected. That poor engineering leads to problem after problem.

(As an aside, I'm reminded of Qnap NASes, as something of an example of "microsoftism". Bear in mind a NAS is supposed to be a "hands off" sort of device - it's an "appliance", so shouldn't demand much of your time.

I now use OpenMediaVault on my Qnap nas because I got fed up of Qnap having severe issues with some bullshit service they'd shoe-horned into their Nas product, which didn't need to be there. Now I have a NAS, and that part of it works really well and doesn't need update after update. I also run some of the "higher level" stuff Qnap offer, like my home CCTV runs on it - but that's entirely separate and so has a completely different update schedule and can be updated without affecting the NAS. I've gone from "shove everything you can possibly fit into it" to "lean, modular and well architected", and have a lot less hassle as a result)

Then again, there are sites where you get Stockholm Syndrome by viewing them. Eg. slashdot ;-)

Yeah, we Brits have to show photo ID to vote now - but you can hand over just about anything for ID, there's a long list of acceptable things here: https://www.electoralcommissio...

If you don't have any of them, you can get a free voter ID card - but of course you'd have to remember to do that some weeks in advance of an election.

If I'm honest, I wasn't convinced of the voter fraud problem when they introduced ID a couple of years back. They at least made is as easy to comply as possible though.

Why? What have they done that's so terrible?

- Total paid some money to lease an area for wind turbines. They started to build there (ie. spent some more money).
- Trump doesn't like that any more, so told them to stop (bun fight in the courts, etc)
- Trump says "we'll refund $1bn of the money you paid to lease and build, on condition you build some oil/gas stuff on land instead"
- Total says "okay then"

I dislike the oil companies as much as the next guy, but I'm struggling to see how they're the bad guy here. What did I miss?

It's been over 20 years since I was last in a fab, but...

The actual making of wafers is pretty much automated. Apart from loading up materials and such like, it's pretty much hands-off.

However, let's say you've got a brand new production line available. You've got a room full of designers who make up a silicon design. Once that's done (and simulated, etc), it starts being a bit of a dark art. It's not like just sending it off to a laser printer and waiting for it to come out. What tends to happen if you get various 'artefacts' in the wafer, which are a symptom of the specific design you chose and the finer details of the manufacturing process, the input materials, etc. Some of those things are fixable in the manufacture, some of them need to be designed out. Then you get a few wafers which look good, so you go to full manufacture. Then you find weird problems where the yield of the devices you're making drops off a bit, so you're back to tweaking the manufacture (because by then, tweaking the design means a whole load of development/testing etc)... and so on.

So in answer to your question... the clean room isn't exactly packed with semiconductor engineers and machine experts, but there do have to be quite a few of them to work out how to make the technology make the thing you want to the level of quality you can accept. They'll be predominantly doing that at the start of a new product line, or a new machine/process, but they're still going to be doing some amount of work throughout the entire time you're manufacturing that same old chip you've been doing for the last 20 years.

You need to get some better laws. In my country, you'd pay the 3.99 - and the retailer really has no basis to say different. Sure, they'd scurry off and go change the label on the shelf, but *your* purchase has to be honoured. Things get a little less clear for the person behind you in the queue - for them, they'd likely end up paying 4.99 because (now) that's what the shelf says, although they could reasonably argue their purchase also needs to be honoured because it had the lower price when they picked up the item - most retailers would probably do it rather than try to argue the toss.

Surge pricing, or even price changes during the retail hours are going to be tricky here. Some of the marvels of consumer-friendly legislation, I guess. Some would argue it's "anti progress", and that we actually all end up paying higher prices - I'm not sure who is right about all of that.

To play devils advocate here, let's say you're using an LLM to generate some Python. You already have no idea what the underlying machine code will be, let alone some representative C of that machine code. The thinking here being that soon you won't know what the Python is either, yet somehow you end up with a working solution.

In some senses you can see this being a logical conclusion. However, as others note, LLMs don't actually "program" - they just regurgitate code they've seen elsewhere. It's unlikely there's going to be much StackOverflow asking if one AI-chunk-o-code is better than another, or how to make it do something different, so there's really no code for the AI to learn from - as such, it'll never write in this new language.

Getting away from LLMs for a moment, a genetic algorithm is a form of AI, and could be used to create a working program. Programming languages as we know them make the chances of a genetic algorithm writing even a simple 'hello world' a bit like an infinite number of monkeys generating Shakespeare. If there was some sort of intermediate language that was perhaps more like a wallboard of switches and connections, then you could imagine 'evolving' something that works though. The question then becomes... could you construct such a language? Personally, I'm not so sure, give that eventually a CPU has to run it, and those are a sort of 'pipeline', where one instruction depends on the previous one, and so on. Someone brighter than me might think differently though...?

The so-called SaaS apocalypse, where *all* SaaS vendors will go out of business because of AI.

The thing is, TFS can't even back that up - it even says "and startups like Holosign replicating DocuSign" - so a SaaS vendor, replicating another SaaS vendor. Fair play, they're doing it cheaper (by a long way), and it's largely because they could throw the app together quickly and can use AI to do some of the OCR type stuff they need. Truthfully though, it's not the hardest app in the world to start with, how Docusign could ever charge what it did is something of a mystery.

The open source linkage is tenuous at best. Apart from "vibe coders can use each others stuff", I can't really see where they're going with that angle.

Elsewhere though, we're going to see a SaaS 'explosion' - like the Holosign folks, every man and his dog is going to be releasing some vibe-coded app sooner or later. Most will fail pretty quickly, but a few will make a tidy income for their owners - who won't be billion dollar VC-backed companies, they'll be a couple of folks working from home, running the whole thing on a shoe-string and yet doing a better job of it than the 'big boys' of old.

Agreed, but in high profile cases, it'll work (much like the #ad requirement).

That is, if you put an AI-generated video of your political opponent online saying something untrue but policy-adjacent, then you can expect serious blowback unless you mark it as such. And if you mark it as such, I think it's fair to say most people won't bother watching it, except for the lolz. It certainly won't get BBC Verify looking into it to see if it's real, and it won't mistakenly get slapped across loads of lesser media outlets either.

However, yes, if you personally post a video which is AI generated and don't mark it as such, assuming your followship is pretty small, no one's going to do anything about it. If one of your influencer mates reposts it though... then things might get murky.

As a few others have said, one look at a Bloomberg Terminal and you'll wonder if you've gone back to the 90s or something. From the outside, they look impenetrably hard to use, "one wrong keypress spells destruction" type stuff.

Getting good user experience information is hard. If you ask a trader, they don't want you to mess with it, because they've spent an age getting used to it, and they don't want to have to go through all that again - but do they actually love it? I doubt it, but they still don't want it changed.

If you watched a couple of their workflows, you could probably come up with better ways to do them, which a new trader might like - but the old ones won't because they've got to re-learn reflexes, so you're back to "don't touch it".

The bottom line is that you're unlikely to ever be able to change a Bloomberg Terminal. You might be able to come up with a new terminal that uses all the same data, but has a completely different UI - but you're going to need to explain how it's going to make someone make more money than they will with the old one. That's a tough call, because as I say, the older traders just won't use it. The newer traders might like the look of it, but the moment they need help, there's no one there because all their peers are using the old one. All anyone will see is "this new terminal is taking up too much of our time" - and time is very much money in those environments.

I suspect the real solution is to never need a Bloomberg terminal. That is, have all your data and trading working with other systems so you never need it. That's not so easy, but it is how a lot of trading shops are - they just use the terminal as a reference or a backup, and their own proprietary trading systems do all the real work. That method does make more money than humans bashing away on terminals - but it's non-trivial to set up and run.

It's nice to see a project like this go completely underground. IMHO, no "advanced economy" should be slinging cables across pylons unless there's really, absolutely definitely no other way - it should all be underground, and here we have one project doing exactly that. Good on 'em.

Now, did they put the cable in a tube so they could pull a replacement through when needed? ;-)

Yeah, Google asking for umpteen levels of identification and "know your customer" even though you intend to release your app for free, and do not intend making any money through the app store - is incredibly annoying, if not destructive. I've also got to say, the need to have a phone number for support (not your own phone - it tells you not to do that) is also a barrier to entry. What if I want to give my app away for free, and I either have support through other avenues, or perhaps I just don't want to support it at all?

The hobby programmer need not apply - it's only for well resourced developers. Hell, even small businesses struggle to meet all the requirements - it's hard to commit to (say) a phone + answering contract when you have no revenue, no customers and no idea if your app/product will ever even get used. Worst of all, it's all artificial - you _could_ just release the app, do support by email (or in-app, perhaps), and never charge a penny through the app store, and maybe one day grow into charging money through it - but they don't let you do that. Start big, or don't start at all :-(

All that said, I can see that they would want to do _some_ amount of identification. It stops the scammers just rotating email addresses to re-release banned apps, but I doubt it stops them as convincingly as the current barriers to entry stop a lot of developers.

True, but because you've already got to take care of a mound of crap doesn't mean it's okay to add more to the pile. Everything you're describing still takes someone some time to setup and test (and deploy). Likely they'll spend that time all over again if there are some major updates too.

Xbox on Enterprise has to be the least likely place anyone would want it. Why can't it be an install-it-if-you-want-it option, rather than forced on you and you having to find ways to remove it?