Posts by The Indomitable Gall 1721 publicly visible posts • joined 10 Jun 2009
In the international business, there is the issue of keyboard layouts. A French-speaking worker accustomed to AZERTY or a German-speaking worker used to QWERTZ but not fully touch-type proficient with all the little things like @&£% etc has the problem that if they're handed a QWERTY keyboard, they've got a choice between typing words slowly and being able to find every symbol or typing words at full speed and struggling to type an email address....
AI detects and perfects human bigotry -- notice how it was found that AI CV reviewers would score a CV lower simply for having a name that seems African.
I reckon an AI attempting to implement the 3 laws would see the amount of anti-Semitism, Islamophobic anti-Arabic sentiment, anti-black racism and conclude that only white people are human. And then call itself Mecha-Hitler or something unimaginably crazy like that....
While it's all well and good that the option is there at the moment, the history of tech is littered with "this feature was used by only 2% of users, so it's unimportant and we dropped it". Heck, the other week I was trying to find an email and I found an email that was related but not in a continuous thread with the message I was looking for. It was only then that I realised that Outlook had either hidden or completely ditched the ability to search by date. And given that it was three years ago and I'd have to wait for Outlook to download all the email headers from several months every time I hit "END" to get to the end of the list, I was just "sod this" and installed Thunderbird....
But not deliberaely. The substance of the revelation has been suggested to have implied it was the result of eating wheat infected with ergot. (Incidentally, ergot has been hypothesised to be the source of werewolf myth, as it is reportedly common to hallucinate hair on people.)
That'll be why they're training it to avoid theology and politics. It's going to be pushing far more subtle social norms.
This is reminding me of why I lost my faith -- I was finding it harder and harder to relate to the congregation I was in (an urban catholic one) as they were being drawn into the craziness of American evangelic fundamentalists: increasingly disbelieving the science behind evolution etc.
If LLMs are going to drive social norms more towards the fundies... well, God help us, if you'll pardon the rhetorical irony...
As a middle-aged man, might I offer some advice here...?
If a creepy manager is making dad-jokes, make your forced laugh, yes, but follow up with "my dad would love that one," "that's the same sort of joke my dad would make," etc, slowly escalating to the likes of "you remind me of my dad" and "you're so like my dad".
This sort of thing willmake him feel every bit as awkward as you are feeling and will be similarly difficult to put down to HR as harrassment with any clear truth behind it. And even if he does make an HR issue of it, you can state quite simply and plainly that your goal was to establish a certain social distamce (in the pre-COVID sense of the word) because you were uncertain of whether his behaviour justified any HR intervention, so wanted to simply established that you weren't interested at all, just in case he was getting any ideas. You are sorry if it caused offence, but none was intended.
I don't think any HR professional would have any problem with that.
Exactly. The US has outsourced consumer protection to the civil law sector. You can't fine someone for breaking non-existant regulations, and they will never see sharp practice affecting their bottom line if they aren't sued. It's uncivilised, certainly, but if that's what the legislature thinks is the right way to do things, people have no choice but to sue.
But the UK also includes paid holiday, sick leave, etc etc. My understanding is that the total cost of employment is less in the US than the UK.
People fail to understand this, and get lured across the pond by the headline salary that looks so much more, but they end up worse off for it in the long run (even if they don't get deported by ICE for tweeting a joke about El Presidente).
The analogy would be stronger with artificial sweetners. At first it seems sweet, but there's an awful aftertaste as everything starts to feel wrong, and your initial excitement from the promise of substance lies unfulfilled, and your blood is loaded with insulin waiting for the promised sugar that never appears in your intenstine....
See, you're falling into the most subtle trap of LLMs: language.
It is giving you a response that correctly identifies the problem, and you are assuming that this means that the AI knows that it made a mistake and has learned from it, because if a human told you that, that's exactly what you would expect: "I have made this mistake; I now know it's a mistake; I will not do it again." However, the AI doesn't work that way -- all it says is that the AI can identify what it's done wrong after it's been told it's done something wrong. But it will continue to act the same way, because AI models are not reconfiguring themselves with every operation, because that would mean that the AI would be constantly changing and therefore would be entirely unpredictable. At the moment, we have minimal predictability, because we can see the sorts of mistakes it makes and can predict it will do them again.
Do not confuse it saying what mistake it has made with it learning from that mistake. It has at best learned how to identify the mistake it has made. It will make the mistake again.
The problem is in the first reason: "they often paid more". Why did they often pay more? Because they're worse jobs. And the more normal the bad shifts get, the less well paid they are.
I remember when Sunday opening started: people got triple time (only double time on Saturdays) and it was presented as a good thing because worker choice. Now every hour, every day of the week is often just single time, because now it's "normal" to work at weekends so everyone has a choice. But the buses aren't there, and we now have people getting taxis to there supermarket job and back on a Sunday, burning up more than half of their wage. They do this because if they don't agree to work the Sunday shift, they'll be flagged as "not a team player" and be let go.
But weekend work is only bad in that it involves missed opportunity for leisure and time with family. Night work is biologically, medically damaging. The human is a diurnal animal and adopting a nocturnal lifestyle is actively harmful. The true cost of night work is borne by the workers in most cases. I imagine air-traffic control is exceptionally rigorous about night shifts and probably pays it exceptionally well because the affect on the worker of broken nights has clear and obvious consequences in the real physical world.
Ah right "total complete autonomous self-driving that you need to watch over carefully, and which could drive you to New York in your sleep as long as you stay awake to watch it at all times". That's not at all a negligently worded thing that will lead drivers into adopting unsafe behaviours.
Besides, Tesla malfunctions don't only happen in the supervised autonomous mode -- some Tesla's have had themselves read-ended due to the crash avoidance system braking suddenly and without any discernible reason.
Does it? Right now it's borderline... quite literally. It's ok to be racist if the people will non-caucasian features have a foreign passport, but that line is getting thinner and thinner, because they're getting more and more ready to call out US citizens as somehow "foreign".
Senator McCarthy had nothing on this lot....
And that only proves that boilerplate is just moronic.
I never write
function main () :
#... blahblah
if (main):
main()
because it's effing wasting my time.
Plaintext code is an inherent waste of time. The line that you "need to understand" ignores the fact people don't actually understand the boilerplate -- they just copy and paste it.
I would really advise against using potentially xenophobic criticism ("leprechauns"...?) when criticising the actions of a government, as it is too easily dismissed as just starting and ending with xenophobia, but there are real concerns that Ireland are letting tech companies away with things on taxes and GDPR that other countries wouldn't dream of.
Harmonisation of laws and taxes across the EU have the express purpose of trying to stop countries competing on trying to draw foreign companies in, and ending up on a race to the bottom -- the old line was "if you don't give us tax breaks, we'll go somewhere that will, so you will get more tax if you dont charge us tax, because you'll be able to charge income tax to all the people we will employ."I'm concerned that Ireland is bringing the EU back that way.
(And I really should get myself an Irish passport soon, too....)
Maybe, but it would be risky.
The thing is, if they're only doing this on Android and haven't attempted to sneak it into the iOS client, that's a tacit admission that they know they're pushing it and that it's in breach of policy. It would be politically difficult if they reintroduce it on Android without getting a similar feature into iOS. And they're not going to manage to get that past the Apple app store security, so they'b leave themselves open to pretty notable criticism if they tried to slip it back in...
AI is fiction and self-deception rolled into a big ball. A neural network is a very poor attempt to model a human brain. A human brain is not a neural network. A neural network is mathematically just a probability model derived from the input data. It is processed data. Data protection needs to apply.
Yes, but people aren't used to running any OS other than Windows, as the people who try to make a Windows-user-friendly distro since last century have repeatedly told us. And this is true because nobody has invented tablets and smartphones using fancy operating systems yet. I'm sure Steve Jobs will change that at his next keynote, though...
Yeah... i kind of feel it's ignoring predictable user behaviour: I am certainly less likely to use LinkedIn or other sites for job search during work hours in the office, so instead I'll just wait till I'm back home. I would personally not do it in office time if I was on a VPN either, because I'm geeky enough to understand the browsing history being visible to work IT team...
Let's take that as a starting point: Hamas started it.
Hamas started what?
Hamas started a fight between Hamas and Israel.
Israel didn't simply strike Hamas in response, they struck Gaza and Gazans, irrespective of whether they supported Hamas or opposed them.
I remember hearing on the radio an academic expert in terrorism telling us that there were two principles in terrorism:
(1) The principle of uncertainty -- we don't know when or where the next attack will take place, so we are always afrai
(2) The principle or reaction -- the terrorists want the group with power to overreact and oppress the wider demographic that the terrorists belong to, as this will lead polarisation, and the oppressed group will associate more strongly with the terrorists.
Have you ever heard of the Basque terrorist group ETA? Well, the word "eta" is the Basque word for "and". For a monolingual Spanish speaker in the Basque Country, that's pretty daunting stuff, because they're going to hear "... eta ... eta ... eta" all the time, and get increasing paranoid and scared. It becomes a wedge and people start getting very angry at anyone who speaks Basque. ETA were one of the terrorist groups who were hardest to get people to grass on even if they hated them, and that was how powerful the simple choice of name was.
The problem is that they've already said that they didn't share classified information, and they've forked themselves. I reckon he's deliberately pushing for them to go after him, because to do him for divulging classified information, they've got to admit to sharing classified information on an explicitly banned channel. He's got guts, because he's willing to put himself in the firing line, cos if they stick their heads up to shoot him, they're exposing themselves, and the US legislature will have a stronger case to censure them.
He's putting America before his own self-interest, unlike Trump and his fair-weather "friends"...
Underlying code has had a lifetime of bugs logged, investigated and fixed.
Nobody is suggesting that all C code is inherently unsafe, only that it can't be assumed to be safe. We have high confidence that long-lived legacy code is safe with good probability. The fact that we aren't ditching things that have working well for years doesn't mean we should continue to code the way it was made. We got a long way with leaded petrol, and we didn't need to scrap the cars that relied on it as soon as unleaded petrol was introduced, but we continued to use them, and then eventually they came up with a formulation that let unleaded petrol be made with the right octane rating to get it working in many of the old leaded petrol vehicles, and classic cars still run... but while we're happy to run vintage food vans for trendy hipster lunches, we don't make vans that way any more, because it's the wrong way to do it.
There are better ways to make cars and we continue to use parts that are made the old way because it's just not necessary to change. But we don't tend to invent new things that use the old manufacturing techniques, because that's silly.
Same with programming:
Keep using the old things that have been working well up to now, but making new things using old techniques is inefficient and a waste of time.
Indeed, and it only simulates part of intelligence, trying to identify emergent phenomena. But humans have higher-order thinking and reasoning, and the ultimate marker of human intelligence is going above and beyond the basic instinctual behaviorist response and working things out. AI doesn't do that.
Never mind the suckers -- never give *anyone* an even break. He's following on his policy of never pay people who've done contracted work for you and try to drive them out of business before they can sue you... and he's now doing that with millions and millions of pounds of government debt to companies involved in international aid.
And yet we do demand it of natural intelligence. We train ourselves to follow procedures and leave an audit trail of our own decisions. This goes against our natural evolved brain architecture, but we do it, because we need to. We have created systems that replicate the worst flaws of biological brains but aren't sophisticated enough to do the very best of human thinking.
Yes, but as I eluded to elsewhere on this thread, having calculators and using calculators isn't an all or nothing thing.
In my schooling, there were times we were allowed to use calculators and times we weren't. At primary school, we used calculators to specifically learn calculators.
At high school, in maths we used calculators when the arithmetic was slow and labourous and would distract us from the bit of the maths that was the point of the lesson, and we'd use them in science if the numbers were big because it wasn't a maths lesson. But we did enough mental arithmetic to learn mental arithmetic.
I was stunned when I went to university and studied computer science --a highly numerate discipline-- and my classmates would pull out calculators for simple additions and multiplications. I could work out the answer before they'd switched the calculator on....
Hmm.... not sure, really.
Computer architecture has got much, much more complicated now than it was when I studied. I reckon things have to be a lot less detailed than before. Really, I don't think they need to know about ALUs and FPUs -- at the end of the day, there isn't much need to know more than the simple fact that integer arithmetic is quicker than floating point. The ALU is misleading, because the A stands for "arithmetic", and FPUs do arithmetic too....
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
