Posts by NerryTutkins 285 publicly visible posts • joined 9 Jan 2018
Starship payload capacity in tonnes is waaaaaaaaayyy too optimistic. Two stages inevitably makes it less efficient, carrying a lot of excess weight to orbit. So it's quite likely to end up as a massive rocket with relatively low capacity.
The talk of 10 or more starship launches to provide fuel for a single moon landing is insane. You cut the price per launch to a fraction and then need 10 times as many launches. The maths suddenly doesn't look as great.
I think they were great cars a few years ago, but have actually got worse.
They removed LIDAR because of cost, even though it performs far better. They removed well proven ultrasonic parking sensors and again use the cameras, which simply doesn't work properly.
I don't really have a problem with the fact they are still using the same body shapes (I think the industry convention of brand new generations every 7 years is kind of unnecessary if the old one still works) but it does help give the impression they are not really keeping up.
5 years ago, they were the only established choice if you wanted an EV that could go further than just local, and could charge faster. But now there are many other options that are better, or cheaper, or maybe in the case of Chinese cars, both.
Musk seems incapable of acknowledging mistakes. Other manufacturers are rolling back to physical controls for many features, because customers prefer that. But Musk seems to plough on, convinced of his own genius on everything and just doubling down even when the evidence is clear that he's wrong. It's the kind of vanity that led to the Cybertruck.
I think his extremist views have certainly not helped, especially considering the EV demographic, but I think Tesla would be struggling regardless now at least in Europe. Not only is there a lot more competition, Tesla have actually made their cars worse by removing working tech for cost reasons.
The US is so divided, that the only real solution is simply the break up of the country.
It would be nice if it happened peacefully, but unlike the democratic EU when any country can leave, the US simply does not allow this.
Civil war is inevitable. Because I do not believe that decent, productive smart people in the big economic powerhouse states like NY and CA that vote blue will meekly accept being governed by fascist billionaires and inbred hilly billies.
All true, but this presupposes that MAGA voters care about pedophiles and the rich and powerful committing sexual offences against kids. I guarantee they do not.
They cared about sexual immorality, until Trump's "grab em by the pussy" and Stormy Daniels exposure, and then private life was suddenly unimportant. They cared about democracy until Jan 6th, and then they supported insurrection. They cared about pedophiles when it was fitting up democrats with ridiculous pizza parlour sex rings, but now that Trump's the top pedo on Epstein island, I guarantee they don't give a toss.
MAGAs aren't motivated by facts, they are motivated by performative cruelty to people they hate. It's simple for Trump to keep serving up misery to those they hate, and they'll happily ignore anything else he does.
When Trump said he could shoot someone in Times Square, he wasn't wrong. He could just as easily have said they could watch him fiddle kids in Times Square. None of them care.
There is zero chance the GOP will lose the house.
In the 2020 election, the only thing that stood between Trump stealing the vote and later clinging on to power was a handful of republican governors and attorney generals who did the right thing, and a VP who refused to go along with Trump's insurrection.
Now, there is no way that Trump will allow his total grip on power to be reduced. He has the supreme court stitched up, all dissent and opposition has been removed from his party and every GOP state, official, congressman and senator, as well as a majority of supreme court judges, are 100% behind anything Trump does.
Trump will fix and steal the election, and at that point the democratic states will have to choose. It will be clear that the democratic system is no longer functioning, and therefore there is no way to peacefully remove the MAGA regime. They either meekly accept the verdict of the activist supreme court, or they reject the USA, form their own confederacy and seize control of their own states.
I upgraded my desktop, which doesn't have TPM 2.0 or a supported processor, using some unapproved "hacks". This was not difficult, didn't require any geeky coding, just downloading a couple of tools to create appropriate installation media on a USB drive that would bypass the various requirements.
So much as I think suing Microsoft is a bit over the top, they really could give people with unsupported hardware a much easier way to keep those machines running that would almost certainly work for most of them. Dumping Win 10 users into either paying for support, or going without security updates, is a stupid idea. And I suspect it will come back to bite Microsoft.
5 years ago when there were few EVs available other than Tesla unless you wanted a small city car with low range and pedestrian performance.
But now the market is awash with EVs that directly compete in terms of price and performance with Tesla, and often with better build quality. Tesla has also screwed up things that cars have done right for years like rain sensing windscreen wipers and parking sensors, because of Musk's cheapskate insistence you can do these tasks with the cameras alone. For the same reason, Teslas dumped lidar and so rivals are way ahead in terms of self-drive and adaptive cruise control that can actually be trusted.
As well as big European brands which do the car side really well, the Chinese are eating Tesla's lunch on the tech and battery side.
For anyone in the market now, there are really no compelling reasons to buy a Tesla. Certainly not in Europe.
I've reported scam ads on Youtube to them using their process, and they simply do nothing. They acknowledge the report, but there is never any follow up, and I continue to receive the same ads.
They are obvious scams.
Over winter it was "this trick the energy companies don't want you to know will heat your whole house for less than 10% of the cost of normal heating" and the ad is basically selling a little fan heater with built in plug that you plug into the wall. It quite obviously only delivers whatever heat corresponds to the electricity it uses. Basic physics. It won't magic heat out of nowhere.
Now its summer, I see very similar ads but now it's "the device that's taking the world by storm, it can cool your house for a fraction of the price of expensive air conditioners"
There is quite obviously no oversight whatsover of the ads, and even when alerted to scams, they do absolutely nothing. The same ads just keep going and keep going.
Can we just take a moment to appreciate the hilarity that Facebook is actually going to court to stop a customer from repeatedly purchasing services which its web site has quite happily sold to them.
Imagine the brains trust at Meta getting to this point. Surely first they thought about implementing some kind of block, or getting their super hot AI to review advertisers? But they decided to go to court rather than fix their own broken process.
The problem is, why would anyone invest billions when the US policies are so erratic and half baked, they could be cancelled tomorrow, or you build your factory and they then decide they don't like foreigners owning successful factories either?
Better to head for the exits, which it appears the stock market investors there have already figured out.
Curiously, the conservatives in the US have extremely liberal views about shagging porn stars while your mail order sex trafficked former hooker illegal immigrant wife is nursing the last of your batch of kids you've spawned with three different wives.
"live and let live"
"peoples private lives are not our concern"
etc.
Nice to know the UK isn't the only country with forelock tugging serfs.
With flip flop Donny, will anyone be stupid enough to invest billions building factories in the US? He'd just as likely change his mind next week, or change his mind and punish you anyway because your business is foreign-owned. Being erratic and unpredictable is a very bad basis to encourage long term investment.
Most companies will sit on their hands, so I would expect investment to collapse in the US. Already there are huge flows of cash from US investors to EU stock markets. That suggests smart money in the US is heading for the exits.
Many of the countries targetted have made clear they have no idea where the US got the reciprocal percentages from. It turns out, they basically made them up. What they appear to have done is taken the trade deficit with a country as a percentage of imports from that country, then divided in two. So it isn't really about tariffs, it is about offsetting trade deficits. Even countries with zero tariffs on US goods get hit if they happen to run a trade surplus with the US because they provide better value, quality, etc.
The question is whether the US economy is large enough to be able to have a trade war with the rest of the world at once. 60 years ago, it probably was. Now, almost certainly not. Rather than move production to the US, in many cases countries will simply move around the US. So EU buys things from Canada it used to get from the US, or from Brazil, etc. And those countries buy stuff from Europe they used to get from the US. It will still cause damage outside of the US, and to some countries more than others. But it mitigates the damage, while the US faces the full cost.
In terms of moving production to the US, would anyone want to commit billions to do so when the regime there seems so unpredictable, especially when the country is likely to be lurched into a recession or even a depression? Imagine committing billions to the US and then finding either that the Trump regime is ejected in 4 years time (assuming they have not stitched up the political inconvenience of elections by then) or that Trump simply flip flops as he has multiple times already and cancelled or delayed tariffs indefinitely. As a company, you'd need to be absolutely certain that the US will maintain these tariffs for the long term (beyond next election), that the US economy will survive well enough that there is still a market there worth serving. Very big iffs.
I visited the UK with my kids and wife a few years ago when they were toddlers. We'd gone to Burger King for lunch as the kids loved junk food.
I arrived back at my parents' place in the car and had got the kids out and sent them in, then got chatting to the neighbour who I'd not seen for many years and who'd known me since I was a kid.
I noticed at times he was looking a bit weirdly at me, but it wasn't until 15 mins later or so when I went inside that my parent's commented they liked my hat. I'd completely forgotten I was wearing a Burger King cardboard crown.
My point was that "it's complicated".
As you say, the reality of gender is that it is far more nuanced that whether you have a cock or a fanny. But that's precisely the point the "two genders" crowd dispute.
And they seem to be the people most vociferously objecting to this particular woman boxer. If they don't think she's a woman, and obviously not a man, then how does that fit with a "two genders" position?
And if they try to square it by saying she's a woman, but with some physiological advantage, then surely the same should be applied to any athlete with freak physiology that benefits them. For example Miguel Indurain and his oversized lungs springs to mind, and Ian Thorpe's massive feet and hands, but I am sure there are many, many others.
I cannot help feeling that a guy with big lungs or big feet doesn't arouse any of their prejudices, but a "butch" woman does.
To be fair, I think the UK system is just as vulnerable to dictatorship as the US.
In the UK, there is no written constitution, and no supermajority in Parliament required for anything. So if you can get 50% + 1 of the seats, you can do virtually anything. Apart from abolish the monarchy, which legally there is nothing to hold to account.
Like the US, the only thing that has preserved democracy has been a general deeply held respect for it, if not by those at the top, at least in enough of those below to make it impossible to overturn. Trump's attempted coup last time was only thwarted by Pence and a couple of Republican state governors being unwilling to go along with his attempt. But anyone who will do the right thing has been purged from the party now.
A UK government could, for example, change the electoral system (even to a profoundly biased one), change the terms of parliament from 5 years to 100 years, if it could get more than half the MPs to vote for it. And first past the post makes a single party having a large majority very common. With Musk now assaulting European democracy and pledging millions to the hard right, this is a real prospect.
Which is why Labour and any centrist politicians really need to consider changes to the electoral system such as proportional representation. It doesn't solve the problem but it makes it much harder to get a massive parliamentary majority with just a third of the votes, as Labour achieved last time out. What benefited them last time, could easily benefit extremists in the future.
I found it curious how the "there are only two genders" crowd also seem adamant that the Algeria woman boxer during the Olympics should not have been able to compete as a woman. If there are only two genders, which is she? She has a vagina, and no penis, and that's how she was born, she's not trans. Surely that makes her a woman, and thus eligible? I mean, she's quite obviously not a man so by a process of elimination, she's a woman.
They are all for simplicity and "common sense", until suddenly it doesn't suit them and then "it's complicated".
Yes, it is. Life is complicated. Which is what people have been trying to explain to you for many years.
When I look at Facebook now, my feed isn't posts by my family members and distant friends.
All I get is post after post from people I don't know, that FB is showing me. And these are either
(a) "Glory to Russia" posts, showing the Kremlin or some pic of Russia with sentiments of "Russia, the best country in the world" or
(b) Sites claiming to be "Historic Britain" or "Past Glory" showing old, or AI generated pics of London or other UK cities, full of white people. Looking at the transparency info, which FB tries to make hard to get to, these sites all originated as innocuous sites several years back, often in completely different countries, which have at some point been changed to their current incarnation. Judging by the volume of Reformtards posting, the purpose seems to be to elicit racist commentary on how much better the UK was before Sadiq Khan let all the dark people in and took over.
It doesn't matter how many times you click "not interested" and "stop showing me this shit", Zuck's shit cannon just keeps sending the crap, relentlessly.
Is Zuck an agent for Putin, or are FB engineers just a total bunch of asshats who can't figure out how to actually give decent recommendations?
Is it really so hard to create a social network where I can choose just to see posts from my family and friends, without having the billionaire owner's shit cannot hosing me down every time I login?
Step one ... get the oligarchs on side, get them to buy up the private media
Step two ... get the government to take over what private media the oligarchs cannot buy
Step three ... crack down on what is left of any private or independent news outlets so dissent is all but eliminated
Step four ... look forward to elections, which will continue, but will be a pointless farce as propaganda overwhelms everything while the government roots out uncooperative judges and guts the power of any political bodies the opposition controls
It seems a little naive to me to think the US will have another presidential election, or at least what would qualify as a meaningfully fair one.
People in the US know that Trump is an insurrectionist, and the coup last time failed only because of a few principled republicans (state governors and the VP) who did not go along with it. Such dissent has been purged, MAGA is in total control now.
Trump is a big admirer of Putin, Orban and Erdogan, all autocrats that came to power democratically and then gradually destroyed the democracy to stay in power. He will not make the same mistake he made last time of failing to prepare for losing, he and his gang will attack independent news and broadcasters, while Musk and Zuckerberg pump out propaganda so there is no meaningful opposition. And in the event even this does not guarantee a win in the next election, they will ensure that control of the courts, and the US president's immunity, will be used to do *whatever it takes* to stay in power.
The only hope for the US is that some states decide to leave. I know they are not allowed and that could lead to civil war, but the alternative is to meekly accept dictatorship and see those states have their own power crushed. Because dissent cannot be tolerated.
As for the rest of the world, democracies need to circle the wagons, because the US billionaires are coming for every other democracy next.
This seems like a shakedown. It seems to be targetted at a single hosting operator, which looks dubious.
I could understand if they announced a commercial service, made the terms clear, had a clear fee structure and then every host who wanted to access the .org resources would have to pay some kind of fee. But picking out a single successful operator looks more like it is manipulating the market for its own benefit (presumably because it offers pricy hosting services that this company competes directly with).
This is the problem with FOS... eventually projects either die through lack of resources, or they succeed in which case the owner has leverage to start extracting money from users.
The good approach might be a clear "premium" option or added services available from the start. The bad approach is to build a huge following then try to blackmail companies successfully exploiting your free software and if that fails, start to use your free project's heft to steer customers to your own competing services.
There is a big difference between criminal trials and civil litigation.
In criminal cases, the case must be proven "beyond reasonable doubt" (i.e. 99% certainty of guilt). In civil litigation, it is typically "on the balance of probabilities", i.e. above 50% certainty.
So being acquitted in a criminal trial, and then losing a civil claim for the same actions later is not unusual. OJ Simpson springs to mind. It does not suggest that the civil action is without merit, or the result flawed.
The premature accidental deaths of two key defendents at almost the same time looks bad, but almost certainly is a coincidence - the freak wind that hit the sailing boat would be impossible to predict or create,
The issue is self-driving cars. What is the point?
"Reduce the workload of the driver" - free up the driver to do what? You still have to sit in the driving seat. You still have to have hands on the wheel. You're still legally obliged to not fall asleep, or tug yourself off, or watch youtube or play with facebook on your phone. You have to sit there, and watch the car drive itself, and be ready to step in at a moment's notice to stop it driving off a cliff, or into the side of a truck, or skittling a motorcyclist.
The self-driving tech might be 99.9% reliable, but when the 0.1% can kill people, it's not enough.
This is being done completely the wrong way around. The driver should be driving the car, and the technology should watch them. Make the driver do what they should be doing anyway, pay attention. But add the technology as an extra safety feature that can intervene if the driver makes a mistake, is going to jump a red light, is about to change lane when a car is approaching, etc. This way, drivers don't rely on it. Think of ABS brakes. Nobody goes around slamming hard on their brakes as a matter of routine, relying on the ABS to slow them down without skidding. People brake just like they always did, but in the *rare* occasions when you slam your brakes on, the system kicking in might get you out of trouble.
Distracted drivers was a bad enough problem, especially in the US, since most drivers learn on an automatic where you can drive one handed thumbing through facebook, unlike in Europe where most drivers pass very intensive tests on manual cars that require two hands and full attention. But introducing even more technology that (despite the lame disclaimers and warnings) only assists distracted drivers to be even more distracted, is a recipe for disaster.
One day, machines will be able to drive cars better than any human. We'rere not there yet, and probably won't be for many years. Until then, these systems do nothing to improve safety, they simply enable lazy distracted people to be even more lazy and distracted.
And I haven't even got into the legal liability issue of whether the lazy distracted driver, or the company that sold them a product which positively encourages lazy distracted behaviour, is responsible.
In my opinion, this driver should be held accountable for the death his stupidity cause, and the company that enabled him via such dangerous technology should be fined heavily, and perhaps held criminally liable too.
I am not sure Rust and .NET are oriented towards the same tasks.
.NET is widely used for web development, and Blazor (which sits on .NET) seems very much the future in terms of where Microsoft is going with that. I very much doubt any of that will end up using Rust.
Microsoft has made it quite clear that odd numbered .NET releases will have relatively short term support, while even ones will be long term releases.
So this is a total non issue. Anyone who developed on .NET 7 would likely be in a team that is constantly developing and upgrading an active product and needed or wanted to integrate new features without waiting for .NET 8. And when .NET 8 came, it would probably have been simple to recompile to that (I found even .NET 6 code easily compiled without changes to .NET 8).
Anyone building something that wants long term support should use .NET 8 and stick with that even when .NET 9 comes out.
Is it really that hard to understand?
Whenever I have rented a car, I've generally been doing significant mileage (else I'd just take an Uber), often in places I am not familiar with. The limited range of EVs, charging time and potential worries about finding charging mean they're less well suited to most rental cases. So it doesn't surprise me Hertz's decision to go big on EVs hasn't worked out.
I am not an EV hater, in fact, I am a happy EV owner. But at home, we generally don't do many long trips, we know where chargers are if we need them, and we generally charge at home where there is a cost saving. Most rentals are business trips, so the main advantage of cost savings isn't going to be a big factor if your company picks up the bill.
British industry was primarily destroyed by bean counters. Sure, the unions probably didn't help. But other countries in Europe that had strong unions such as Germany seemed to keep a successful manufacturing sector.
I worked in Germany for a year in the 90s, for an engineering company (I am a mechanical engineer by degree, but drifted into software development) and the contrast with the UK was quite notable. In the UK, company management tended to be people in suits with voices getting posher the closer to the top you got. Very often they graduated in marketing or business, because of course engineers should stay in grubby workshops doing real work, not making management decisions. In germany, engineering companies were run by engineers.
While the UK car industry had accountants choosing components and making design and investment decisions based on £££, most german companies had more interest in the engineering. Working first hand in germany, I could see that the guys running the company (who were graduate engineers) did things properly because "this is the way things should be done, and we're competing on quality, not price".
Various foreign car companies seemed to have no problem being quite successful in the UK (at least prior to brexit), which was largely down to the way they were managed, not the unions or the workers.
This is absolutely true.
It's like financial advisers, stock traders and bankers. Let's say there are 5000 in the country. Even if we assume it's just pure chance how they perform, 500 will be in the top 10% in a year. The next year, 50 of those 500 will be in the top 10% again. In the third year, 5 of those will be in the top 10% again. Now this is just statistics, it could be more, or less. But based on probability, there will be a few in the top 10% for these three years.
Now the industry lauds them as gods, and they probably believe it themselves. One year would be luck, but three years in a row cannot be luck, surely? They must be special and have special skills. Their employers hike their salaries, bestow obscene bonuses and give them even more assets to manage, and freedom to take even more risks. And so when the inevitable happens the following year, it goes from being a manageable decline to being a massive, economy-threatening catastrophe because they've bet the farm on someone continuing to keep rolling sixes.
The amazing thing is that banks and the financial industry, which absolutely should have the mathematical background to understand very simple chance, get caught out again and again and again. They legally have to put "past performance does not guarantee future success, your funds are at risk" but never take it on board themselves.
So there are two different things called Copilot.
Then there is Visual Studio, and Visual Studio Code, which are two different things.
Then there is .NET, and the legacy version which is now referred to as .NET Framework, which are two different things.
I heard Satya Nadella has two daughters and they're both called Susan.
It is indeed mind-boggling.
One of the common arguments by moon landing conspiricists is that technology doesn't disappear or go backwards, so the fact we cannot at present go to the moon (and have not since the 1970s) means it must have been faked.
Presumably the fact you could supposedly travel the atlantic at mach2 in the 1970s but now in the 2020s you cannot even do it at mach1 means they think concorde is actually CGI. And that your grandparents were in on the deception, like 10s of 1000s who worked on the moon landings.
There was no hysteria, the safety concerns with the AZ jab were well-founded, despite the UK's initial chest-thumping brexity attacks on the various EU health bodies that first raised concerns.
The UK also removed the AZ jab as an option for younger recipients, and then conducted all its booster campaigns with mRNA vaccines such as the Biontech/Pfizer jab. The UK essentially accepted the risks and moved to safer vaccines, though this caused delays in the UK vaccination programme. It may have had a better chance of getting more supplies of the Biontech jab, had it not selfishly acted to block AZ vaccine shipments to the EU early on via a preference contract, which later backfired spectacularly having burned all goodwill with the EU when it need their vaccine.
The AZ vaccine most definitely saved many more lives than it may have cost, because it was primarily given initially to the most at-risk groups, in particular the elderly. The UK rightly accepted (but tried to downplay) the safety risks of the AZ vaccine, in particular for younger recipients, and moved over to the Biontech jab primarily, which had proved safer overall once there was significant data available from mass vaccinations.
This is how science should work. It is sad that UK politicians in particular exploited vaccination as a political benefit of brexit, which made them reluctant to adjust strategy when the risks of the AZ vaccine emerged and also poisoned the well when it needed EU vaccines to complete its own programme.
The claim that somehow verifying the browser is legit would stop bots or other automation from cheating on games or doing other nefarious deeds on an automated basis doesn't really stack up.
Selenium makes it relatively easy to build automation that can control chrome and firefox via an API, those would surely show up as legit, unmodded browsers. Although Selenium is oriented towards automated testing of web systems, I've worked on projects where clients needed to scrape data from various sites (with permission) that did not have APIs and tried to frustrate scraping by having javascript and fingerprints etc. and so needed real browsers to get at the data, rather than just calling the form targets and parsing the response. Selenium has been around for a long time too, and the browser makers specifically put support for it in, so they certainly know this too.
The proposal looks more like trying to crack down on all the webkit/blink clones to keep control of the market and stop smaller start-ups getting a slice. It solves a problem that I am not sure exists, with a solution that doesn't actually solve the problem.
If I have a company expense card, and go out and spend 2000 quid on a slap up meal with wine at a fancy restaurant the day before I leave the company, and the company objects and considers this an abuse of what the card was intended for, they can't sue the restaurant for the money back. That's not how it works. They may have a case against me, depending on what guidance I had been given about the use of the card.
Even if we suppose that the bill was unjustified, twitter accepted it and paid it. So Musk should be going after the former management for abuse of company finances, not the law firm. I would imagine they have a fair amount of cash to hand, given the fact they gave Musk a bath he won't forget in forcing him to buy for 44 billion.
This simulation sounds like it was programmed by the work experience guy's younger brother.
If you create a reward system of points and then intervene to stop it scoring points, it seems entirely reasonable it decided to remove the communications and the operator if they are effectively reducing its score.
This doesn't illustrate a failure of AI as such, it illustrates a failure of those implementing it to put in basic controls and create appropriate rules for the AI to operate under. Quite obviously, it should be told the rules include not harming friendly troops or equipment, or that will score minus one million points for any such violations.
As with chatGPT, the special sauce is not the algorithm, it's the training and the prompt.
I have used Chat GPT a lot.
I have had some very positive interactions with it, for example, code suggestions that didn't quite work, but approached things in a different way to what I'd been doing that subsequently achieved what I wanted.
But also completely false information, followed by another go and another go in which it then gave different false information.
But I would not dismiss it based on this. It is very early days, and it has mastered basic communication excellently. I am sure with suitable training data and rules in place, it would prove an excellent way to automatie simple customer inquires, for example. In time, it may become more reliable in more complex tasks.
We're somehow expecting it to be god-level intelligence when is perhaps the first generation of AI that can converse at a human level. A few years ago, if you'd built a machine that could have a conversation like this, pass the Turing test, and handle routine human interaction, it would be seen as an epoch-defining milestone.
Even if this AI is only at the level of dealing with counter duties at McDonalds or giving passenger advice at an airport, that is still a phenomenal achievement that will have massive implications for many jobs and society as a whole.
The only difference I found with being human is how easy it was to admit a mistake, apologize and accept whatever truth I told it. Even humans who accepted daft stuff without any evidence ("stolen election", "cat litter trays for schoolkids who identify as cats" etc.) are generally unwilling to be put right and will just double down on the unsubstantiated info.
I can see this interaction being used to poision AI too, just like Google bombing exploited Google search suggestions so when you type a name or company, it would make outrageous suggestions that would strongly imply very negative things.
I am very impressed by ChatGPT and have genuinely got some useful information from it that allowed me to solve a coding problem.
However, I have just asked it about an open source project which I am lead developer on. I asked it who the leader developer was.
First, it gave me a name I have never heard of. So I said I had never heard of that person, and they were not the lead.
It apologized, then gave me another name I have never heard of. I told it again, I have no idea who this person is, and they are not lead developer on this project.
So it apologized again and gave me a third name I have never heard of. This time, I corrected it, told it that I have never heard of that person, and the lead developer was me. It said thank you and that this would help to have correct information in the future.
The project is mainly developed through my company, but it never mentioned the company name, nor figured out that as a director of that company listed in publicly acessible records at Companies House, I might be lead developer.
Chat GPT is impressive in its ability to talk like a human, but it is worrying how willing it is to spout utter bullshit with apparent conviction and then so easily correct itself and accept whatever information you give to it. If this is how it works, it is never going to be a reliable source of information, though it may still have utility in making suggestions on coding for humans to test, or doing basic customer service interaction based on a very limited set of training data.
According to the flat earth community ("do your own research"), the moon landings were faked because it's not possible for humans to get to space, let alone the moon. They frequently cite the lack of present tech to take people to the moon as evidence ("we're supposed to believe they could travel to the moon in the 60s, but 50+ years later, we no longer can").
I like to point out to them that Concorde therefore must be fake. Because supposedly in the 1980s, you could fly across the Atlantic at Mach 2, but now in the 2020s, you cannot even do it at Mach 1.
"Yes. Because it's what 90% of my users are familiar with And what 90% of their websites integrate with."
This sounds very much like the rational given in the late 90s and 2000s for using Internet Explorer.
And since web standards weren't really a thing back then like they are now, it was probably a more compelling argument back then.
I've also formed very similar views on testing it a few times with quite diverse questions ranging from immigration law where I live (my wife is a lawyer here and was impressed at its answers), to programming issues.
It is really astoundingly good at situations where you are after facts or curated information, rather than opinions (which is what the moral questions really are).
If you want information, Google can give you simple things directly, but for more complex things, it just presents a list of web site links where the information might reside, and you have to view those pages and find it. What ChatGPT does really well is give you the information directly.
It is not perfect. I asked it whether it could contact law enforcement if a user confessed to it they'd committed a serious crime, as well as telling me it could not, it also recommended calling 911 if I was aware of a serious crime, although the emergency number is 112 here. Clearly with many questions, the answer will depend very much on where you live, and I was surprised it did not identify which country users are in from an IP address in order to tailor things accordingly.
But none of this takes away from how phenomenal it is and I am sure it will improve greatly with time.
Most modern fighters are inherently unstable and can only be flown thanks to fly-by-wire. Machines have reactions and response times no human can match. I am pretty sure that suitable automation could handle crosswinds and wind shear way better than any human pilot could, if equipped with the right sensors and programming.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
