Posts by breakfast 1504 publicly visible posts • joined 24 May 2007
Having a camera is great, but I would miss a lot of pictures if I had to have my real camera with me the whole time. My phone catches lots of things I would otherwise miss. Not to mention that a lot of photos are just useful - showing someone what a problem is so they can give you a quote on fixing it.
I feel like Mastodon would be a pretty good home for the small subset of Reg community members who don't snort huffily at the very concept of social media, being social at all, accessing the internet with anything newer than a dialup modem, and computers smaller than an IBM mainframe. The "complicated" nature of having to decide which server(s) to sign up to has resulted in a fairly tech-savvy community and if you search around there are plenty of funny and interesting people on there and interesting conversations.
In the post-twitter world it might be the social media I have most fun with.
I have only heard the opposite - people I know who update Google about impassable routes either get no response and no change (and have to keep turning back truck drivers trying to get around impossible bends or squeeze down bridleways) or get a change for a few weeks that is subsequently reverted without explanation.
Given that you're not the driver in question, it seems as though your experience driving a different vehicle in a different country might not be applicable here. I'm in a slightly different part of the country where driving 30-35 on many of the single-track roads is likely to get somebody (or at least some livestock) killed. Roads are built different in different places and American vehicles are insane.
Also looking at the photo in the story, if you were approaching at night the way the road continues the other side could lead you to think that the road simply dipped down to the bridge below the range of your headlights. On a road with no warning signs or barriers of any kind (this is the bit that is incomprehensible to me!) why would you immediately assume that you were driving towards a collapsed bridge?
The Muskwit tried to do stock manipulation on Twitter with his offer and then realised he had made a terrible mistake when he had to choose between buying it at the price he had offered and going to jail. Now he's trying to find a way to weasel out of the situation and it's possible he thinks he can line up some kind of bankruptcy deal that will make the problem go away for him, so he's trying to come up with the most deliberately destructive policies possible.
Of course, that may be doing him too much credit, we have established he is so thick the spoon stands up and that he is undergoing some kind of billionaire-brained breakdown, so he might just be that incompetent.
A real advantage of these stories; if only there were some that took place past 2000 and referred to technology that still exists...
On one hand I want to hear about people tying their docker containers in knots, on the other I have a solid "who me" story that I'm wondering about sending in because it feels a bit recent.
Another interesting place where Amazon is selling AI-generated books is apparently guides to mushroom foraging. People who have looked into this suggest that some of them contain advice that is likely to get people killed.
It will be interesting to see who is considered liable when this happens.
When I used Linux as my main desktop OS I found I couldn't avoid fiddling with it if I wanted to get anything working.
Eventually I realised Linux was free as in "free to spend all my free time trying to get basic things working on my Linux desktop" and switched back to Windows. I still complain about it a lot and I still run my machine in dual boot but on a day to day basis Windows just works with far more software than Linux does. Particularly with things like music software, simply being confident that VSTs will work consistently means I can confidently buy them, which opened a lot of doors for me.
I hope that we'll end up with a model that allows desktop software to work well under Linux, but I don't know that it will ever happen.
Counterpoint: Microsoft don't care about those of us who want to use Windows on end devices. We're a small part of the market and simply not interesting to them. If I was selling software that needed to run on a local machine - if I was an Adobe or a Cubase or whoever - I would be looking at how I could get my tools working on Linux because sooner or later Microsoft are going to kill my model of locally installed software running locally.
Everything about these stories shows why we can't trust billionaires with anything. As someone observed back when Twitter existed: "In terms of cognitive impairment [being a billionaire is] probably like being kicked in the head by a horse every day."
Congratulations to Cliff Biffle on having exactly the kind of name I come up with when I'm GMing and the players ask the random shopkeeper what his name is so I have to come up with something on the spot. Big "background NPC who is about to be adopted by the party" energy.
A really neat formulation I saw recently was "if a black person has to spend an hour a day blocking nazis to make your site usable while a nazi does not have to block anyone, then your site favours nazis."
Non-moderation is every bit as much of a choice as moderation. Who you choose to favour is very telling.
IR35 is kind of bullshit but it would be funny to see the Tories becoming the "tax our most loyal voters twice" party if it wasn't a problem I also have to deal with every time I'm considering a new contract*. It's complicated, stupid and it works for no-one.
If I was pushing for political power I think "make tax understandable" would be a pretty good element to one's campaign.
* Hypocritical? Absolutely, but I never voted for the leopards eating people's faces party. Seeing people who did get their faces eaten is schadenfreudelicious.
Only now do they tell us the cloud can be affected if there are too many clouds. Fortunately that's the only way things can go wrong. Except if things get unplugged or switched off and on wrong or there's a problem with a policy update or someone else's cloud goes down and somehow takes a critical piece of your infrastructure with it.
Honestly sometimes it feels like the cloud is just somebody else's computer.
"Prone to generating false information" certainly doesn't feel like something you want in your command and control chain.
There's probably a fun sci-fi story in an army that manoeuvres in specific ways designed to trigger the weakness in command and control AI systems by manipulating what is reported about them.
If someone makes a TV adaptation of a book they have sort out some kind of licensing arrangement under the expectation that everyone is going to make money off the deal. Why should scraping it to train AI be any different? You're using an authors work to create a commercial product that makes money for you.
If you don't want to license it, then don't use it- at the very least authors should be able to say "I don't want my work to be used for this purpose" but like soup ingredients, it seems as though once it's in an LLM you're not going to be able to get it out.
Could but may well not end up federated with a lot of Mastodon instances because by and large the community there does not have any trust at all for Meta. They'll have to work very hard to demonstrate that they have something to offer and aren't just planning some further embrace and extend bullshit, which is what the Fediverse community are bracing for.
Twitter handles a huge amount of data and I expect there is some really clever engineering going on behind the scenes, but you've hit on an important truth here. Perhaps the most important thing that Lone Skum did not understand was that he was buying a community and he needed to operate it in terms of maintaining that community.
Instead he acted like he had a technology product and proceeded to chase the users away by making it progressively more obnoxious and harder to use, destroying its value in the process. The thing that made Twitter worthwhile was that for a long time everyone was there. I doubt there will be another platform like it after this fragmentation.
Twitter's management had been walking the tightrope of maintaining that userbase for a long time. I think had he not bought them out, they might have been failing by now anyway - being the poster children for "grow first, monetise later" starts to get tough when the interest rates on your loans are rising fast and you still haven't got around to making significant amounts of money.
Not at all, quiet quitting is sticking to your paid hours, doing the work that is requested of you but no more, basically working to rule. The bosses get so rabid over people handing over their labour for free that they categorise limiting work to what is in your contract a kind of quitting, while they would never dream of paying you more than the amount in your contract.
The more power somebody wants over others, the more antagonistic they find it when people maintain firm boundaries.
You're clearly very fortunate, but you were also able to retire early and comfortably, which implies you're maybe a generation or two ahead of many of us here.
Suffice to say the market has not moved in the direction of treating people better than it did when you were in the workplace and your experience may no longer be representative.
Don't get me wrong, I'd love to work somewhere that offered those kinds of benefits, but if they still exist now they are few and far between. More of a tech unicorn than a profitable start-up, for sure.
"The irony is that in taking this approach, the quiet quitters are unlikely to achieve the pay rises and promotions they so desire." I don't know if our reporter here has had many jobs, but I can guarantee that working extra hours will almost never earn you a pay rise or promotion either, but now you've wasted half your life working extra hard for no benefit.
If you're working outside the hours you're paid for then you're doing your job as a hobby.
I read a very interesting article on this yesterday (The LLMentalist Effect) that suggests that part of the wonder around LLMs is because we have accidentally trained them to communicate like cold readers or stage psychics. The way their statements are ranked for accuracy in training can have the outcome of favouring the same type of Barnum Statements and statistical guesses that make it easy for the people using it (a self-selecting audience, just like the psychics enjoy) to trick themselves into thinking there is intelligence there.
It's a really interesting article, well worth a read.
Scientists will never get those definitions until they start listening to philosophers, who've been working on the same questions for centuries.
In fairness, there is a lot of crossover between cognitive science and philosophy - people like Dennett work with both - but a lot of the engineers working on AI end up ignoring the cognitive scientists as well.
My understanding is that much of this work is along the lines of going through thousands of pictures and labelling which of them contain a bird. Without that step I don't see how a model could train itself to identify birds because it wouldn't be able to tie anything it was evaluating to any human-meaningful data.
There's a good reason for this.
Imagine if you will that you are an experienced expert in the field of AI, well-read, ethical, imaginative and thoughtful. Now imagine someone asking you to work on murder-drones. You're going to rightly tell them to GTFO.
Unfortunately, they're going to keep making the murder drones, but they only have access to people who are not ethical/well-read/imaginative/thoughtful and the results end up being like this.
To my mind the most likely AI apocalypse is one where we have autonomous weaponised drones and some future Bezos type billionaire tells them to execute anybody who looks like they might be in a union.
I don't believe he thinks that, I believe he thinks that after making his ridiculous offer for it, he couldn't wiggle out without drawing a lot of attention from the SEC. That could have ended up with a sad little Lone Skum in a prison cell somewhere and although that would be hilarious for the rest of us and almost certainly work wonders for SpaceX and Tesla, he believed that it was worth spending a truly implausible money to avoid that. Some of his backers saw a great opportunity to get rid of Twitter as a force for global information sharing by supporting him and here we are.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2023
