Posts by Howard Sway 1822 publicly visible posts • joined 12 Oct 2020
The cheapest Tesla comes in at $39,000. So if you only want one to run around town in, it would be cheaper to buy yourself one of these taxis. You'd also be able to go to sleep on the way to work in it. And get drunk on the way home.
Has the BOFH really fallen for the AI hype? Is he not yet aware that such requests often dish out inaccurate and different answers each time you ask? This is the first episode where the boss sounds, dare I say it, sensible by wanting a simple, unchanging and accurate solution to his problem. OK, he's wanting to use AI to write the program, which just moves the AI inaccuracy problem elsewhere, but his worry about the AI going away is quite plausible, and even if it doesn't get that far, the massive subscription increases coming down the line as AI peddlers desperately try to recoup their huge costs are going to make writing that little program far cheaper than a 10 year AI subscription.
That's what it says above the forecast on the BBC weather page. But the forecast is now right. So it was presented as right when it was wrong, and now they say it's wrong when it's right, which is of course wrong. Right?
Copilot Meeting Summary
Action Items
I can see it all now : one day thousands of programmers go online to order pizzas from work, and find themselves in a crappy version of "The Forbin Project" which ends with them all starving to death after the AI fraud detection system declines their orders and disables their door passes, leaving them all locked in the office.
"In its 2022/23 financial year, Vodafone's operating profit was approximately 14.3 billion euros. This figure represents an increase of nearly nine billion euros if compared to the previous year"
The special pleading when they have more than enough money to upgrade their networks is pathetic. The idea that investing the money to improve their infrastructure will get more customers who pay more for faster services seems not to have been entertained - just keep milking the massive profits from the old stuff for a few more years, take home huge bonuses in the boardroom and let some other poor bugger sort it all out once you've left.
When this happens, the excuse to use is that it's best to leave straight away to beat the rush hour crowds and give you time to write up your report of the work done in the airport whilst everything is still fresh in the memory. The fact that the airport has a bar and you plan on spending the extra time there is best not mentioned.
Is this true? .I have been finding the output of all the LLMs I've experimented with frequently creates more work than it solves, is often useless and contains hard to spot errors that look plausible, but would be disastrous if left in. That's just with single prompts.
Letting some AI agent go off and start assigning itself more tasks based on the output from the first prompt is even madder. Before you know it, you'll have a customer database full of cake recipes, your default font will be Comic Sans and in a week's time you'll discover that it ordered 25 trombones for you off ebay.
The "technical community" for me consists of people who understand technology and whose experience leads them to be cautious with their enthusiasm until they have properly considered the downsides of something and tried it to see if it matches the hype. The person who discovered the problems in this article by experimenting and discovering them is a good example. The people who are saying "woah, the new shiny can do everything!" are generally part of either the salesman community or the gullible idiot community.
No, stick to names, but bruise their egos. DictatorPawn, DisposableLackey and BitterVirgin would be quite hard to reclaim as a badge of honour. As far as I'm concerned they could be as offensive as they like for these losers.
Most critical security issues occur at the OS or OS-adjacent (eg drivers) level. Rather than spreading blame equally across the industry, it's there that drastic changes of attitude and a rebuild from the ground up needs to occur, with 1 company we all know the name of in particular refusing to fix itself, as it favours new shiny features over thinking "what whould a robust, safe OS look like?". Right now is the time to start doing this : computers are fast enough now that complete isolation of processes and layered memory access that is monitored and guarded, at the expense of a slight performance decrease will not cause any pain for users. Bare metal memory and processor access needs to be abandoned for all but the lowest level, simplest, and most highly guarded parts of the OS. A file system where everything's in one directory finally needs to be abandoned in favour of proper sandboxing and transactionality. But, no chance, so we're left floundering about with constant introduction of new vulnerabilities and patches probably forever.
No, you did admit wrongdoing because you paid the fine for it. If you'd refused to pay the fine and been taken to court you would have been saying you did nothing wrong, but paying the fine is an admission that you have no reasonable defence against the accusation, and therefore an admission of wrongdoing.
This may be true to some extent, but it's not entirely a tale of being exploited if $11 billion of other people's money ends up in your account and you were knowingly part of the con for quite a while is it? She could have decided to go to the authorities before the authorities came for her.
Which is the heart of the problem with the idea that everybody will want one. It's like selling cars with retractable wings because somebody might work out a way to make them fly in the future. All the companies who've invested fortunes in AI are desperate to get that money back, even though it has no use for the average person on a local PC, and the fact that the general reaction is "I don't want it" should be a warning to those planning to make us all pay for it anyway that some companies will undercut them on price and outcompete them by not offering it.
Unsurprising when your plan is to use something miraculous that can do everyone's jobs without bothering to find out whether it actually works well enough to do so.
IBM could be the first company where everybody loses their job to AI though - but only because the company's gone bust due to their stupid bet that it could do technical jobs better than a human.
Well it's pretty useless then. Air skates (a mini hovercraft pallet bed) let you move a washing machine around a warehouse with almost no effort. This looks like they have created clever stability control for something that will not make any economic sense to use.
The thing is, Microsoft has quite successfully spent the past few years getting application workloads pushed into "the cloud". So, if all your data and workloads are there, what business processing needs to be done on local PCs? Surely they should be becoming more of a simpler, cheaper client machine for the corporate cloud if the last few years of cloud hype have any truth to them?
It's the old client / server, thin client / thick client, local app vs web app debate yet again, in a new dress, but this time with very few, if any applications where "client side AI" might be needed. And it only seems like a few weeks since they were pushing "cloud PCs". Has that stopped now, as they've realised that they were basically reinventing cheapo Chromebooks for office use?
I'm surprised that they never copied the British trick of renaming nuclear power facilities in the aftermath of headline grabbing melty / explodey incidents. If they'd changed its name to "Rainbowville", nobody would now be batting an eyelid.
Windows originally was an app, that you had to run from DOS, so maybe they're going back to their roots.
If you access another computer using the Windows app, can you then use the windows app on that computer to access the one you logged into it from and continually spawn remote sessions on each other this way until the computers slow to a crawl? If so, you can then persuade your boss that you really do need a new state of the art machine, or failing that be allowed to move to a sensible OS.
Just finalising my updated profile : I've now been the president of Switzerland, the first man on the moon, the inventor of the hovercraft and head of unprincipled user exploitation at LinkedIn. And these are just a few of the many more things I intend to appear as in our chatbot future. I see it as a social responsibility to make the output of these things as inaccurate as possible if they're going to just freely help themselves to everything they can find about everybody.
Why? If your ultimate ambition is for LLMs to do all your programming for you then it won't matter whether it uses functional or imperative languages, or which ones it uses : they all get compiled into machine code eventually, so why can't the LLM just cut out the conceptualising of multitudinous programming languages and go straight to producing machine code?
Anyway, away from fantasy island where we can all easily dream of machines that magically do anything for us, here are the first lines of a program that a LLM made for me at the weekend :
#include
#include
#include
#include
#include
#include
I'm still pondering why it felt the need to #include nothing precisely 6 times...
Let's mandate this constant recording and reporting for billionaire tech bros too. Everything they do 24 hours a day, with an "AI" deciding whether they've done anything wrong. Just to watch how suddenly less keen on the idea they become.
And there I was thinking that his "joke" offering to impregnate Taylor Swift was going to be the stupidest thing he did this week............
He's lucky he's so rich and famous - otherwise his week might have been a journey from arrested development to simply arrested.
Well, it's fortunate that none of our jobs require "thinking" isn't it?
I have been experimenting with LLMs to see what they can produce, and I have to say that none of them are even close to producing real world applications for complex business domains of the kind I've spent much of my career writing custom apps for. This is because they're not trained at all on business processes for individual businesses, which are mostly highly specific and also not publicly available for LLM training.
Amusingly, I got one to generate a high frequency traiding bot for currency dealing, but it was just the basic structure for the app, and had a comment saying "put your trading logic here". It's at this point that all us experienced programmers who've been replaced can expect to get the phone calls asking if we'd be available to do a bit of work, and can start quoting hourly rates in the thousands.
But surely Pat had implemented backups on the old Exchange / AD server, and could have just restored from them, then done the migration again, because he kept notes of each migration step?
Not doing a backup, and risking a complete loss of all mails when the disk failed seems a much bigger screw up than pressing the delete key a few times.
How much more seriously can you work on the idea "I would like to put a tiny little house on the moon"? Did he spend a decade agonising about which shade of red to use? Chimney or no chimney?
Frankly, it's a boring idea. I would have sent a model Moonbase Alpha complete with Eagle transporter.
I just asked it an ambiguous apple question .It knew if was ambiguous, and then gave a completely wrong answer :
If I have 5 apples and take two away how many apples do I have?
If you have 5 apples and take 2 away, you would have 2 apples remaining. However, if you are referring to the apples you took away, you would have 2 apples in your possession. So, you have 2 apples that you took away.
But if I then ask it
If I have 5 apples and take 2 away how many apples do I have?
If you have 5 apples and take 2 away, you would have 2 apples in your possession (the ones you took away). You would still have 3 apples left from the original 5. So, you have 2 apples that you took away.
So, it can't work out that 2 and two are the same thing.
Just in case what? In case your competition is insane enough to do this and drive themselves out of business?
Here are a few more "ideas" you might want to patent to stop the competition getting them first :
* Mandatory in-car AR glasses that turn every other car on the road into mobile hamburgers if the car hears anybody say "I'm hungry"
* Full self driving that takes you to advertiser's shops, rather than where you want to go
* Robotic arm above driver's seat that tattoos adverts on your face whilst you're driving
It shouldn't do. This brings back distinct memories of a company I knew of a couple of decades ago who decided that they'd have to rewrite all their successful C++ applications in Java in order to keep up with the new cool kids in their new web application world. Hey, the languages looked almost the same, but Java was "memory safe", so had to be even better! How hard could it be?
It turned out to be disastrous. Features their programmers were used to were missing from the new language, and the whole application design paradigm needed to be rethought from scratch. Add in the fact that the rewritten application was of course the first project undertaken in the new language, with all the mistakes and bad design decisions that entailed, and they were left with an error plagued, badly performing mess of an application that failed in the marketplace, as did the company soon after.
Well, Google must have allocated a fair amount of its own money for its programmers to learn Rust and do the conversion. But they're proposing that everybody else spend their own time and money to rewrite their applications to do exactly the same as they do now. As small companies and individual companies are generally not overflowing with spare billions the way Google is, it's quite an ask, even when ignoring the risk of introducing new bugs during the rewrite.
The problem is that most Windows software ever built is for x86 series processors, so running them through emulation cannot ever be anywhere near as fast as running them on bare metal. Especially when you're using RISC chips to emulate CISC chips - not a very intelligent or efficient architecture.
Mind you, that's an academic point when the emulation can't even run all programs correctly. There is a big chance that these things could be a total flop in the market, as they'll still be competing against native x86 machines.
This is quite simple passenger information. Addr = 67 means that you have to turn to page 67 of Michael Portillo's book of British railway journeys to find where the train is going. The brackets inform you that journey time will be 43 hours. As for the dot matrix info boards on the train, how else are they going to inform you that a trolley service is available for tea, coffee, sandwiches, crisps, etc, etc? Apart from the 10 minute tannoy announcement listing every single item available of course.
Because the old farts have spent a few decades encountering and solving all the problems that your new shiny has yet to face.
The main benefit of using C in the kernel is that it lets you get really close to the machine - which is a good thing when you have to deal with so much weirdly designed hardware and get the right bits moving to and from it. The code to do this can be pretty ugly, but you can at least get things working this way. The proposition that rust is better for such low level work is yet to be proven in enough concrete side by side examples for me yet.
The organisations he's suing managed to get a brief story about adverts next to neo-nazi shit on his platform circulating for a day or two, which he felt was damaging for his company. By trying to play judicial whack-a-mole to silence his critics, he ensures that every time he sues somebody, the story gets another run out in the worldwide press, and he reinforces the image of himself as a thin skinned rich bully who's so allergic to criticism that his desired image as a free speech champion becomes laughable.
The accusation that Media Matters "manufactured" the content is a joke when it was the code on his company's web servers that served it up. Whether or not a "typical" user sees that or not is irrelevant. The only way he can ever stop stories about his platform hosting neo-nazis is for his platform to stop hosting neo-nazis (who, as a history book could tell him, weren't really big free speech kind of guys anyway).
Of course he doesn't : it's other people's money he's burning. But burning it is all he's doing. LLMs are not a way to general artificial intelligence because intelligence is not founded on language, language is a product of intelligence.
However the huge money being wasted on this is money that could be far better spent on actually useful applications of technology, rather than furthering the hype driven self interest of people like him.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
