Posts by Pete 2 3413 publicly visible posts • joined 10 Jun 2009
> Hollywood can use generative AI to create scripts or stories, but human writers asked to work on them will be paid as if they worked on any other gig.
So the scripts will be generated by AI, but then a writer will be asked to tweak it.
Doesn't that mean that the amount of time they will be paid for - even at the new, higher, rates - will be much less than if they had spent time writing the script themselves?
> Backporting one is a whole different thing and (off the cuff guesstimate) about three orders of magnitude harder
Since you are guesstimating that implies you have not done this.
I have backported plenty of stuff in my time. Some of it is tricky but certainly not three orders of magnitude (you know that means 1000 times?) harder.
Much of the stuff requires nothing more than applying the supplied patches and maybe adding some #ifdefs where appropriate. That would be a reasonable interview task to ask of someone applying for a kernel supporter post.
> I think you *DRAMATICALLY* underestimate the complexity of the task here.
No. See this as the sort of process that a properly maintained kernel source implementation looks like / requires.
And if an outfit doesn't have well organised and documented code repositories and processes, then any candidate worth their salt would spot that and they would end the interview.
It is reasonable to expect that outfits of any size that rely on Linux kernels and apps to have the ability to build their releases from source. Without that they simply aren't in control of their own products.
.
So a basic interview question would be to have a candidate add a fix to an old kernel. It wouldn't need to be a trick question with multiple obscure (are there any other types) of dependencies. Just to demonstrate an ability to use developer tools.
> A third that lost connectivity were doing so at least once a month, and a quarter moaned it continued for three hours or more a week
It needs to be distinguished whether the disconnections occurred between a person's router and the world, or between their router and their device.
Merely saying. "the internet's down" is far too blunt a statement. Especially as local failures hold the potential for them to be under the control of the user, or at least due to the equipment they chose, or the way they use it.
> If your law school exam can be passed by ChatGPT, your law school exam is broken
Not really. While a new sucker client may well have their first meeting with a partner or senior partner, most of the actual grunt-work will be done by interns or unqualified staff. Only with someone qualified taking a cursory look for obvious errors once in a while. And using that "oversight" as justification for the multiple-£100's per hour that the senior charges, for all the hours the junior put in.
[ voice of being that "sucker" speaking here. £200 p.h. when the individual who did the work couldn't even calculate the final award payout]
If a generative AI can shake up that cosy little arrangement and inject some semblance of value-for-money into the profession, then I'm all for it.
There are already automated systems that will appeal parking fines - ones that were issued by someone else's computer, so to have one AI contesting the work of another: isn't that how they get trained in the first place?
> The device will barrel towards Earth and reenter the atmosphere at 0842 MDT (1542 UTC), reaching 27,650 miles per hour
> On Wednesday, leaders from NASA working with the US military performed the final dress rehearsal for the return mission, dropping a dummy sample capsule from an aircraft
It would be impressive if that aircraft managed to "drop" the test capsule at a speed of 27,650 miles per hour (44,500 kph for the other 95% of the world, or 1621844838 double-decker buses per fortnight for readers here). However, I suppose the old adage of never test for a condition you don't know how to handle applies. If the test capsule did fail, there's not a lot the americans could do to prevent the same thing from happening to the real thing.
> fully autonomous flight AI has steered a drone through a racetrack faster than human pilots
Presumably a feat that will soon be repeated in many other sports where speed and therefore reaction times are of the essence. Just as soon as the computers and their power supplies can be shrunk down to sensible sizes - rather than the half-ton of batteries that electric cars need, in order to attain any reasonable range.
Should we hail or mourn the coming obsolescence of motor racing, in all its forms? Aerobatics, with machines able to withstand higher G's? And the obvious retirement of "Top Gun" pilots for the same reason.
Before there was cheap Chinese crap, there was cheap Japanese crap.
They started out slavishly copying western products. But with the corners cut and prices cut even more.
Then they worked out that improving the quality would save them money: returns are expensive to deal with.
Finally, they started slavishly copying designs the west hadn't made!
> 5G has failed to deliver on its promise
Oh, I don't know. It has successfully convinced a load of people to change their phone for no better reason than an increment in a number.
Even better, it has got all the world's telcos to spend billions on new and (it seems) unnecessary infrastructure.
Oh, you wanted features?
> some tool along that chain may accept and follow any ANSI escape sequences included in that input stream, so if an attacker can manage to get some carefully crafted codes embedded in a log file – such as in a profile name or some submitted feedback – you could end up with a mangled or manipulated view of your IT situation. We can imagine some buffer overflow-style bugs could be exploited, too, if present.
All this amounts to is someone saying that if an unspecified piece of software contains an unspecified bug, it *might* be possible for someone, somewhere, to do something they shouldn't.
Which pretty much sums up the entire raison d'etre of software security. But without actually helping at all.
> Hollywood movies have been telling you, 'Don't do it.' And now everyone's doing it," he said. "I'm just like, 'Have you not seen the movies where they kill us all?'
Have you seen Hollywood's early sci-fi movies?
Now look at their "early" AI movies and compare just how accurate silent versions of Jules Verne was with actual spaceflight. Using that as a baseline, we can say that nothing Hollywood is telling us about AI is remotely close to how the real-world future will be.
> Our analysis shows that 52 percent of ChatGPT answers are incorrect and 77 percent are verbose
Without knowing how that compares to the human-supplied answers, it is impossible to form a rational opinion.
My personal experience of seeking help on forums is that many responses are wrong. Many more are answering an entirely different question while the rest are either passive-aggressive replies, arguing that the question is wrong, showing off or disagreeing with what others have posted.
> the actors you never hear about, earn less than $26,000 a year
And as such, those people might identify as "actors" because that sounds sexy and artistic 'n' all. But it wont be the job they do that pays the rent, buys their food or keeps the lights on.
At best it will be a gig they called in for a few times a year. After several (dozen) try-outs which result in "we'll call you"
> Exactly how the air gap will be implemented isn't clear ... It doesn't matter. It won't work
There is an old (older than the internet) saying along the lines of
Those who say something is impossible should stay out of the way of those making it happen
Which could be a corollary of Arthur Clarke's observation about elderly scientists.
> midrange handsets like the A54 deserve a 3.5mm headphone port
Yes. I changed my phone earlier this year. I specifically excluded every model that failed to have a headphone socket.
I have some very nice wired earbuds that I am quite attached to. Especially when they are in my ears! I can think of no reason to make them obsolete. Especially when all the wireless earbuds are both acoustically worse and much less convenient to use.
It is a well worn part of the geopolitical dance that when one country expels another's diplomats for spying, that the accused country will also expel spies embassy staff from the first.
The USA therefore has no reason to act surprised when China reacts in kind to moves made against it. I am sure this was entirely expected and factored in to the pros-cons when the first anti-Chinese sanctions were proposed.
Does anyone seriously believe that the real motivation (on both sides) is anything but protectionism?
> check that the named researcher actually wrote it.
I'm not sure how that would work in practice. Consider a case where a "professor X" submits a work for publication.
The journal's editor phones the professor and asks. They say "yes, I wrote it". How much credibility does that add?
But we must also consider two other issues.
First, most published work has many authors. Often with the most prominently named authors being the most senior and frequently the least involved. Would we expect a single writing style from several contributors?
Second is the issue of false positives. A paper that a person did actually write getting flagged as AI. How should that be dealt with.
Finally, I'd say that the onus should be on the establishment the work was performed at to police their own staff. So that their reputation would suffer if their staff were found to be submitting work that was not as it appeared.
> none have the software support and the community which the RPi has
And luckily for the RPi people none of the other hobby-targetting SBC manufacturers tumbled that the hardware is nothing without usable software.
Many SBCs have offered better features, faster speeds and lower prices than Pi's since well before COVID. However their software often consisted of an out-of-date o/s and precious little ability to access the I/o without complex and badly documented configuration commands.
Though if all that people want is a media player, pi-hole or other purely software solution then the Pi people may well find that those potential customers have already found other boards that satisfy their needs.
> do you want your medical advice coming from a secret source whose origin cannot be revealed?
Isn't that what is commonly called "experience"?
Some years ago a large computer consultancy set about automating a lot of processes. I has an external person following me around for a week, taking note of how I worked and in particular how I debugged operational issues. Each time I solved a problem, this person would ask me the process I had used to reach a solution. Most times the reply would come down "I have 25 years of experience, it looked like something I'd seen before".
Which wasn't very helpful - though admittedly I felt no obligation to be helpful ;) -, but was the truth.
Yes, this does seem to be an example of the double standards (bias against?) employed in many fields.
Many articles that talk about autonomous vehicles give the impression that nothing less than perfection is acceptable. Yet the standard of AI driving is already (measured as accidents per 100,000 mile / km) better than the average for police drivers.
This seem to be a confidence issue rather than one about actual abilities. Maybe what's needed is some blind comparisons: doctors' diagnoses vs. machine,and see where the truth lies.
> Mean Time To Innocence (MTTI) – the amount of time it takes to prove that a problem isn't your fault.
I generally find that as a first approximation, blaming a problem on the network usually works. If they somehow manage to wriggle out of it, then laying the blame at the door of whichever software back-ends the problem application is a good second bet.
Once NASA had succeeded in getting people to the Moon and back, they were left without any way to capture the public imagination and therefore budget.
It seems that this time, they have learned that time is money and that the more time something takes, the more money they get. Or on personal levels of their staff, the longer their people can spin-out a project, the longer they will keep their jobs.
> making the task of software support easier for Redmond.
Here's a radical idea. Why don't MS make support easier by writing an operating system without bugs in it?
And for parts that are outside of their control, such as third-party drivers, design it such that their own code is robust enough to survive poor design.
Maybe a more (or less, who can say) interesting application of this sort of tech, would be to create some rather unlikely duets or collaborations between musicians who would never have been able to work together. Whether because one or both were utter arseholes, or because one died before the other was born.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2023
