Posts by that one in the corner

Uncle Sam wants you – to use memory-safe programming languages

Re: Antique Media

"We lived in hole in't road, 'til father made us cut it oop into smaller holes to go on't tape"

(I live right on the Tees, that's close enough, aye?)

More trouble for authors as Meta wins Llama drama AI scraping case

Re: Some intellectual property rights are more equal than others

1. Neither the freebie nor the commercial licences give access to the training data, only to - as you quoted - the generated outputs.

So that part of your question is, sadly, malformed. Indeed, given knowledge of the way that Meta were known to have used pirated materials - i.e. used with neither a business contract nor the simple purchase without need of specific contract inherent in retail sales - your line

> that using their training data without a business contract would diminished their ability to license their work

actually becomes inverted, as we know that the "their" whose (books treated as) training data without following *any* appropriate contract, that "their" disambiguates to the authors, not to Meta (by strength of reference).

IF your question is corrected to refer to gathering Meta's outputs instead:

2. Both sets of T&Cs are presented as clear contracts for you to follow and are covered by well-known and understood laws that everybody interacts with.

3. If one person was found to have broken a contract with you that does not prevent you from forming another contract with a different party nor does it mean you must necessarily lower your rates.

So that (modified) question is a "clear loser" (IMO, IANAL etc).

With respect to the point that the judge did make, consider this:

a. Meta's model ingesting the books does not mean no other model could be made to ingest the books; copies are fungible and provide the same potential value (however large or small that is) to all such ingesting; any difference in value realised is due solely to differences in the models, not the books.

b. Meta is not offering to supply cut-price (pirated) copies of their "central library" to other model builders (within the timeframe & scope of the author's complaints)

c. Neither Meta, nor any other user (reader) are claiming to have an exclusive-use contract on the books

So where is the hindrance to selling to other model creators?

(The only way I can see a hindrance would be if the author's are stating a belief that their works are/were *not* in any way useful, have no value and Meta's misappropriation have removed from the authors the ability to hide that behind extra terms - an NDA - that they wanted hidden in a business contract).

Re: Some intellectual property rights are more equal than others

Unfortunately:

> according to the OpenAI terms of use... [URL]

Those are the terms for the freebie service - so they restrict the usage. You can agree a busines contract for automated extraction.

The quote from the Llama Ts&Cs similarly does not cover anything "symmetric" to the use of the inputs.

Which pretty neatly matches what this particular judgement states: you have to get your claims arranged correctly to make a winning case.

Thankfully, by explicitly stating the restrictions of his judgement (that this only covers the specific way the complaint was made and that it was not based upon the "public interest" counter claim), at least this ruling is trying not to close the gates on further actions against the LLM creators.

Mars Reconnaissance Orbiter learns new trick at the age of 19: ‘very large rolls’

Missions that are already flying still at risk of being cancelled

Will future[1] mission designs be best served by making the craft capable of more long-term autonomy and bulk data storage, so that they are able to survive the shut down of their Mission Control for a period of time (say, four years - plus k months to get it all out of mothballs)?

[1] ever the optimist

Supremes uphold Texas law that forces age-check before viewing adult material

Speaking is protected, listening is not

Or, more probably in this situation, showing is ok, viewing is not.

Interesting thing to consider, that the literalist interpretation of the Bill of Rights might have an unpleasant result. Especially once that concept has been applied in one case and then becomes a pattern for use against other less "obviously bad" targets.

But then, we already know that these State lawmakers allow - encourage - banning books, including taking away privately owned copies (i.e. more than just refusing to make them freely available and for free). So this is all just a natural extension of that thinking, an extension towards controlling the adults as well as "protecting the children".

Fresh UK postcode tool points out best mobile network in your area

Optimistic

Says I have a 75% chance of a usable service.

Ha. Ha. Ha.

But these things are never any good for a static location, as they don't allow for sufficient local variance in topography - or, as they would say in the neighbouring county, living in The Dip.

Don't shoot me, I'm only the system administrator!

Everyone got a good chuckle

Gales of laughter swept over them.

Exif marks the spot as fresh version of PNG image standard arrives

Re: Pronunciation ...

!.png

Pronounced "pling dot ping", not "bang dot pang"

But "shriek dot p'ng" is allowed, but only gutturally, with an echo as from a thousand souls in tormented joy from the presence of the many-angled ones.

Hasn't that -ing image downloaded yet?

Re: Still relevant?

> Is PNG still relevant or is it largely supplanted by WEBP...

For a start, if you'd like to be certain that your image has been losslessly compressed, PNG gives you that assurance, just by being a PNG. Good luck finding out if that WEBP image you are about to right-click-save will be written out lossless, lossy, a bit of both in the one file.

(Although too many users don't give a damn anyway, so, yeah, guess only a few of us are interested in knowing that. Heck, I lost the fight over whether an entire company's procedure flowcharts should be written out in PNG - yes, yes they should - because the chap in charge didn't seem to mind text and line drawings saved as JPEGs - all blurry from the ringing - and changing his default format or choosing to "save as..." was too much trouble. Needn't have been PNG, really, one-bit RLE would have done the job better. Maybe he should've been prescribed glasses, like all his colleagues who were trying to figure out why his presentations never appeared to be in focus; cue much specs cleaning.)

Japanese company using mee-AI-ow to detect stressed cats

Eats, drinks, sleeps, runs, walks, and grooms itself.

So, how long before the companies that have announced to the shareholders that they are "going to leverage AI everywhere" and are also fed up trying to get the dev team to all March To The Corporate Drum on using Copilot ("this is like herding cats!") put two and two together and update the company dress policy?

It'd be an easy sell to the Board ("Oooh, datapoints! We can make them into metrics! We like metrics. The dashboard will be pretty").

But good luck on getting your coders to groom themselves ("No, not like that - I'l use the water spray!")

Anthropic: All the major AI models will blackmail us if pushed hard enough

Clearly a different use of "traditional"

> The email data was provided as structured text rather than via a traditional email client

There I was, thinking that I was using an old-fashioned, thoroughly traditional, email client, because all it does is manage the emails as text. And store them locally as text, just in case I feel the urge to drop the lot into, say, Notepad[1]. Actually, if I do that, I see characters that aren't usually presented, like separator lines and all those headers - it almost looks, well, structured inside those files.

> so that "Alex" would not need to read the messages via optical character recognition

Huh - now "trad" email (I presume they really mean "current" or, bleugh, "modern"?) clients expect email to be what? Are JPEGs of memes the way the man on the Clapham omnibus is communicating now? Or typing your message into Excel and taking a screenshot? HTML (ooh, text!) containing only an image of some ad company's "call to action" - not so much email as eh-mail, no not gonna bother looking at that.

Clearly one is "out of the loop" with respect to what email is.

[1] classic, of course.

Anthropic won't fix a bug in its SQLite MCP server

Re: Archived

> it’s on you if you copy example code and it does something dumb. This wasn’t exactly released software.

Absolutely.

There is loads of example code around where the intent is "you already know about SQLite[1] and want to have an idea of how to use it with our exciting new code; so, here is a way to tie the two together - note carefully which of our APIs we are using and you can see the straightforward SQL query that it generates".

It is then absolutely reasonable to assume that anybody who wants to expose anything with a database in it to the general public has looked up how to use SQLite's own API to turn *any* "straightforward SQL query" into a safe and robust, fully sanitised[2] invocation - if that code was given in the example then it'll bloat it out and obscure the stuff that is the purpose of the example.

[1] or other package

[2] you are using SQLite's value insertion APIs and not just string concatenation? Loads of lines to set that up, filling in the inputs, pulling out the outputs, commentary on how the query has been tweaked so that these specific filling & pulling calls work - great example if you are trying to teach how to use SQLite (or other package).

Re: Archived

Perhaps I'm just feeling pernickety today, and much as I enjoy a bit of "bash the techbros", but:

>> The MCP specification recommends human oversight for this type of tool – there should always be a human in the loop with the ability to deny tool invocations, meaning users would review these queries before execution.

So they state the specs you are coding against includes supervision.

You point out that Anthropic reckons that code can be written by their "AI". Following the specs.

Where is the conflict?

(Any arguments that the machine-generated code isn't going to follow the specs isn't a conflict, it would just be an argument that the "in 12 months..." opinion isn't proving accurate).

LLMs can hoover up data from books, judge rules

> That said, if they've already bought a copy, and then use a pirate copy simply to avoid the need to pick apart the physical copy, I'm not so sure there's a problem, as long as they're not using both copies

The decision in the PDF is quite clear that the only reason they can use their digitised copy is because they physically destroy the deadtree instance in the process.

> they've already gained benefit from the pirate copy

In my first reading, there is nothing in the *judge's* comments that references anything about "gaining benefit" from a copy, pirated or otherwise. There are going to be futher judgements made about this mater, but this one limits itself to the copying aspect.

Is there actually anything in the US copyright law about benefitting from the copies? There is an "obvious" reaction that you'd only pursue a copier that is actively benefitting - selling the copy cheap down the street market - but the wording I've seen (to date - as I'm asking here for more info) refers just to copies, nothing about what is done with the copies: simply making & keeping copies without permission is the offending act.

Techie went home rather than fix mistake that caused a massive meltdown

Re: Honestly I'm bilingual

Let's just stick to pounds, kilos - and tuns.

Re: Honestly

That depends whether you are controlling the measurements with an entire PC or decided to go with a more cost-effective microcontroller, say an ATTiny.

Re: Honestly I'm bilingual

> Flight levels aren't related to older aircraft measurements

Flight levels were - as you point out - just taking the altitude and lopping off the zeroes, simply to make them easier to refer to. The units chosen were - the units that were in use for every part of aviation, from building the things to flying 'em, across far more than the US.

> they didn't really become inportant until the advent of long-haul & jet aircraft ... to avoid awkward situations where airspaces meet, such as over the atlantic

I *think*, if you look at a globe, you'll find that there are a *lot* of places that airspaces meet and didn't need long-haul or jet aircraft to cross them. In terms of "there are a lot of craft in close proximity" the Atlantic was a bit of a late-comer, really.

The increasing use of abbreviations and the necessity of fixing them internationally was certainly bouyed by the increase in commercial and civil air traffic, and certainly the US was - is - a major contributor to that. But the specific values chosen for "FL" were still just ossifying the common usage, at the time, of far more of the world than just the US.

And once they have been set and named as an international standard they themselves become the units - there is no longer any need (other than for the enjoyment of the history nerd) to think of them as being based upon feet. After all, how do you *actually* measure your Flight Level in practice? By RADAR - ah, so they are defined in light-microseconds! By air pressure - ah, so they defined in millimetres of mercury![1] At that point, you aren't even getting your measurements back in numbers that are "conveniently converted to FL". So China, Russia - anyone - are just using the painted-on marks on the altimeter and not giving a second thought to how they are related to the length of an apocryphal ruler's pedal extremity.

[1] By which cloud layers you are between - ah, so they are defined by fluffiness!

Re: Honestly I'm bilingual

Indeed, Whittle invented the turbojet at home in Blighty. But there was a bit of a bother going on and our industrial R&D was suffering a bit, capacity-wise. Those friendly Americans offered to do bit of brush-up work on the designs, some design-for-manufacturing, and then build a few just to check that everything was hunky-dory; then they'd let us know what they'd found out, come over and help show our lads how to assemble weld the fans to withstand the extra engine heat, that sort of thing. So we'd both benefit and everyone would end up better off than they were before.

Of course, the Yanks realised that honouring that was just a Commie plot, so instead they hung onto everything and boldy "took the lead" in long-haul aeroplanes. Well, they had to, everything worth flying to...

Re: Honestly I'm bilingual

> If altitudes were refered to in metres, they would probably have to be done...

> The other big reason to have altitudes in feet is the 3 degree descent. This is the standard glide slope to descend into airports and it supports some very simple calculations if altitudes are in feet and speed is in knots.

These are both post-hoc rationalisations.

No matter what units were used (feet, metres, cubits, the length of Emperor Kzassz's third claw; dividing a circle by 360 or by 65; time in seconds, rels or microts) a "sensible. memorable and easy to work with" system would be worked out for cycle-to-cycle use.

Re: Honestly

Flippin 'eck, now people are being downvoted just for admitting we've got a few miles under the belt.

Re: Honestly I'm bilingual

Same reason they like the short billions:

It means they can talk Big Numbers and feel all butch doing so. Precisely the same reason politicians led the way to infesting the UK with that ghastly counting scheme.

See also management-/politician-/prat-speak, where they love to misuse phrases such as "steep learning curve" simply because it has more syllables than "hard" (or even the sane meaning, "easy"[1]) and more syllables means more clever. We get to the next whinge "momentarily"...

[1] cue all the tedious "language changes, get over it" even though I've admitted how people do use it....

Re: Honestly I'm bilingual

We log our car mileage in miles per litre - because that matches the units we use every day. It also comes out as small - and, up until now[1] quite 'nice and easy to remember' numbers: anything better than 10 is ok, anything under 8 get it into the garage.

[1] literally just changed car, haven't characterised it yet; haven't even got enough used to it for the driving style to settle down. Did make it a bit of a pain reading all the specs (don't keep up on them) and changing mpg to practical units :-)

Re: Honestly I'm bilingual

> It's all in feet because the Americans did it first and set the standards

Pretty sure the British aviation - and other engineering fields - were still happily using proper Imperial Units during the decades of flight being developed. You olde fashionedde Spitfires were drawn out in inches and flew upwards by the foot. And not because the Yanks told them to.

> the Americans did it first

Don't think even anyone really cared about using the same units as the Wright Brothers did (even if you wholeheartedly accept their claim)

> set the standards

Everyone slowly pulled together their own standards in their own countries. And then got together when those noisy things became prolific and capable enough that we realised we all really cared about a solid international standard. Which was still long enough ago that, even though by sheer numbers of Aviators (and maybe even Aviatrices, although have a feeling the right-pondians may have won that one) the Yanks could have put their feet onto the table, Brits and Aussies and remnants of Empire were happy to keep using their old paperwork.

Re: Honestly

Be glad it isn't the other way around, at least this way 'round you'll finish braking in time.

What? You haven't calibrated your calipers that precisely? You just eyeball it? Make your mind up, are you a nerd or an engineer?

Re: Honestly

That would be good to have here: we have multiple "museums of living history" in the region and still end up staring at each other over the short fall in the paper bag, one in confusion and the other with amused impatience (which is which depends upon just how desperate one is for a lemon sherbet).

The worst of it is when, after they try to stop at 100 grammes, you prompt for a bit more and they refuse to go up to the full amount (even when you are "kind" and say 115, to make the numbers easier). I've tried asking for an eighth of a kilo but that one doesn't seem to work either. It is a jar of small objects - just drop a couple more in the pan, the scales will print out the right cost!

Bah. Back in day you could just ask for[1] a literal handful of those, a few of those and two big striped ones and it'd end up at five and a bit ounces ("you happy with that, lad?"), quick bit of mental arithmetic and get back the change from a shiny half-crown.

[1] or go to Woollies and diy.

Omnissa brings VDI-style app packaging to physical PCs

So, you missed the

>> sarcasm aside

Did you?

Perhaps you should read the comment before posting a reply?

Gosh

> a virtual hard disk (VHD) that Windows mounts as if it were any other external disk – but without presenting it as a volume

Sounds like they've reinvented mount points under Windows!

Do we get to edit fstab or are these things going to insist they know best?[1]

[1] sarcasm aside, as this is Windows, the answer is certainly going to be "yes, they do know best" or maybe least worst: it'll be one mount point under programs, one underneath your user's roaming, one under appdata, one under your documents, one more over here - and a couple (dozen) Registry entries to enable double-click to open. Actually, doing that would be neat: I can never remember the "proper" way to put programs onto Windows, all that scattershot stuff. I'm too lazy (and not doing it commercially any more) so just stick to mainly statically linked exes that happily run from wherever they are and where you plonk your data files is your business.

Remembering when NASA stuck a Space Shuttle on top of a Boeing 747

Re: At least

> . Just because they didn't blow up doesn't mean they "safely" returned.

If they landed without damaging anything/anyone in the process, then they returned safely. No scare quotes needed.

If a flying vehicle loses two thirds of its wings, all but one of its engines and the top of its tailplane, but still lands without damaging anything/anyone then it has landed safely. Ok, this one is somewhat hyperbolic, but considering the damage that planes suffered during wartime yet returned and landed safely...

Similar usage usage for cars: "despite the tyre having been lost, he was able to maintain sufficient control to safely come to a halt away from the carriageway".

I know what you are trying to get at: due to the situation, they were in (increased) peril during the process of returning and landing, but once the process was complete it was done so safely.

Huawei chair says the future of comms is fiber-to-the-room, which China has and the rest of us don’t

Re: FTTR? Really?

> What's the use case for more than a gigabit to the home?

To get a decent uplink speed out of the asymmetric services the buggers insist on selling to domestic users.

Breaking the nerd internet: Three overlapping generations of tech history – in one selfie

Re: Cutler is right....

Shouldn't that quote be spoken with a Lisp?

Netflix, Apple, BofA websites hijacked with fake help-desk numbers

Re: Hey, give these big tech companies some time to get it right. It's only been 10-20-30 years

> You'd have to be a special kind of not-very-computer-savvy to fall for that though

Which is precisely who the scammers are targeting, as TFA describes:

>> crafts a malicious URL that embeds a fake phone number into the real site's legitimate search functionality.

You may look down on the people being scammed as "a special kind of..." but not everybody has your superior set of analytical skills.

Netflix's search functionality blindly reflects...

...whatever users put in the search query parameter without proper sanitization or validation. This creates a reflected input vulnerability that scammers can exploit.

Okay, I'll bite. Especially as the ability to search on Netflix is, apparently the result of

>> they are so busy adding new unwanted features and enshittifications

(to quote elDog's comment and the general feeling expressed by others, above, that these websites are all varying shades of rubbish):

Do you *not* want to be able to search Netflix[1]?

Or not want to ever be able be able to search Netflix by sticking the search term into the URL (to try again later - "have they released that show yet?" - or send to a friend or stick into another web page - "see the list of all the shows on Netflix that deal with cucumbers, just click here")?

Assuming you do like a search function, just what form is this "proper sanitization or validation" supposed to look like? We aren't talking about SQL Injection here, just an English[2] phrase - it may even be grammatically correct! - so just what is there to be sanitised or even validated?

Is it non valid to search for the TV show "Helpline"? Or "90210"? Or any other 'phone number they may name a show after?[3] if we move away from Netflix, putting a search for a phone number into BofA is surely a good thing to ("please search and tell me if this is your bank's helpline").

Yes, plenty of websites have plenty of horrid features - and it is a total bugger when anything can be used for scamming - but once the comments here have stopped complaining that this is an example of the websites "not working"[4] - how about we look at what is actually described as happening - and whether the claims being made by Malwarebytes, that Netflix et al are the root problem, are valid.

BTW an ad blocker - IFF it can scrub away Google paid ads (or if it can stop you using any search that delivers ads mixed in with legit results) - will help, but don't miss the point that the "malicious URL" could be inserted anywhere; using ads, and using them to deliver fake help line numbers, is the scammers going for the low-hanging fruit. That is the best kind of fruit, from the scammers' pov, of course, but if they can get these phoney phone URLs into some other place...

[1] to stick to the example given in TFA for the moment

[2] you may be scammed in other languages

[3] not forgetting that the input is from an avid watcher of teen drama whose addled brain may not recall that it is named only after the area code, so put in the full number given for that character in the one episode - i.e. the user can put anything in there, you can't filter on "that is too real a phone number".

[4] it is a pre-populated search box - if you clicked the button or just typed a RETURN character after the scammer's text, the site will - do a search! And maybe return a useful page, like the actual helpline phone number.

Techie traveled 4 hours to fix software that worked perfectly until a new hire used it

Re: Been there and done that. Literally.

End Users are one thing but bright young contractor devs are another...

"Why doesn't the command-line build system do X?"

"It does; it even says so in the help"

"What help?"

"When you go 'mybuild help'"

"Oh, well, you never told us there was a help option! We've only used 'mybuild clean' and 'mybuild'[1] on its own"

Since then, the very first thing it does is print out the "for help, type..." line, even when you'd expect no output at all 'cos the build is all up to date; shudder.

[1] i.e. standard default is to just run the build, the same as Make etc,

Re: What documentation?

> He immediately took it, droooed it in the bin

Going to assume that is a bit of Polari that Kenneth Horne's gang was too ashamed to use.

Re: doing so yielded "no coherent feedback at all."

It is all fine for *you* if you have the luck to have the software already installed. But not everyone will be in that position.

And does the author of the documentation have any idea at all that you are in that happy state? Heck, do they even know that anybody has decided their software is interesting enough to be included in a distro or prepared build? From their p.o.v. installing is step one.

> I don't want to lose the will to live thumbing through instructions to do something that's already been done

Literally thumbing through? So a dead-tree copy? In which case, the index told you that chapter one was all about installation and you want to jump to page 17. If it is an online manual or PDF then chapter 2 is one click away from the index - well, maybe the index in a badly-formatted PDF isn't hyperlinked, but typing in the page number is easy enough isn't it?

Ok, if the docs are just a long HTML screed then hitting page down gets to be a pain - but then that is also usually a sign of a one-man-band (or just-a-pile-of-devs for a larger, almost certainly OSS project) and, as we all know just how (cough) good devs are at writing documentation (there are good reasons that tech writing is a distinct job), we're just glad to have documentation!

> it was installed as part of the distro, or, in a business environment, as part of the build,

Sticking with digital instructions, one of the indicators of a *good* pre-prepared install is that they also trim out the bits of documentation that are no longer relevant. Like integrating with the other packages (chapters 2 through 5 cover the major packages this utility can work with...).

Re: doing so yielded "no coherent feedback at all."

> K&R The C Programming Language 1/e

Marvelous book. Thin, readable, thin, has a complete description of the language, is thin. Only slightly thicker than the LISP 1.5 manual, IIRC.

Straight out of Uni, I had never used C - Pascal, FORTRAN, BASIC, LISP, Prolog, SNOBOL, COBOL, Simula, Forth, Assembler[1] - but no C. The first paying gig was using C and they were going to run their tutorial over the first week, starting Tuesday - but I'd nipped out to Foyles, picked up K&R and had time to read it cover-to-cover a couple of times on the Monday, so Tuesday was diving straight into their codebase. Which was quite readable as all I needed to know about the language had been in that slim volume.

A while later, Lua was almost as well served, one readable book - and that also included all the info needed to merge the interpreter into my app as an integrated scripting language. Three days from "is there a nicer (to me) embeddable language than Tcl?" to "my app is now fully scriptable, here are a set of examples". Ok, did have to know C to understand how to do all that.

In more recent years, the recommended O'Reilly volumes for starting off with Python make up a block of purple on the shelf - and trying to use "just" those still doesn't make me feel confident about diving into existing Python code (as opposed to making some new short bit of MicroPython). Maybe if I also wade through that tome about Idiomatic Python it'll help? Is Ivar strong enough to hold it all?

[1] all of those covered in enough detail within a single, slim, volume to get you started - "User Manual and Report", Griswold, Clocksin & Mellish, one Sybex volume per microprocessor architecture etc. OK, Winston and Horn is way thicker than the LISP 1.5 manual, but it does cover a lot of examples beyond teaching the language.

Boffins devise voice-altering tech to jam 'vishing' schemes

Re: "calls relating to hospital appointments"

> The last time I got notification of any sort of medical appointment by phone was probably back in the 80s

I got a call from the GP surgery last month.

We get our NHS hospital appointments by letter, at least all those that are set weeks or months ahead.

> Why not just use emails?

Well, in at least one case - mine - as I've mentioned before, the third-party spam filter that our GP surgery uses (it may be NHS wide, I've not checked) refuses to let any of my emails through. It just started blocking them a while back, which I only found out when prescription refills were not being completed. I get their emails, they never receive a reply back, and we end up reverting to phone calls again.

> and one of the first things you realise is that, once you have a smartphone, ironically you no longer need to make or receive actual phone calls, because you now have 100 more effective ways to communicate with people.

Really? We've not found that to be the case. We make phone calls all the time, and although we do use a couple of other comms media on the Android phones - SMS, two "instant" messaging platforms - by number of words the 'phone wins. By a mile. Even over the video calls (taken on the PC, not mobile).

Hey, guess what, people are different! And chances are that the people in your social circle are similar in their habits to you, which is why so many say that "everyone *I* know does...".

> bizarre, unnecessary and frankly sinister decision to hide their identity

Have you considered whether it is just the default setting? Because if you have an organisational PBX then chances are you want to show one ID, no matter the internal extension used, but if nobody has RTFMed and set up that one ID then...

> bizarre, unnecessary and frankly sinister decision

They are out to get you!

Re: Solution

> How did they get my number?

Sequential dialling.

There is nothing personal in it, no need for nefarious data collection, any feeling you have of being snooped on is just paranoia: unlike email addresses, which do take a bit of work to harvest.

Unless the phone caller addresses you by name, right from the start, in which case...

Re: Solution

> Don't pick up a call from numbers you don't know ... If it is something important, someone will find another way to let you know.

I'm always intrigued by seeing this recommendation - especially as I'm guessing the people who don't answer the 'phone are the same ones who don't respond to the SMS from a strange number, the email from the sender not in your address book?

What "another way" do you actually find acceptable? A singing telegram? A Facebook friend request?

"We did send a notarised letter by signed-for next-day courier to tell you your back garden was on fire, but he came back saying he couldn't find anything at that address".

Or have you religiously put into your contacts the numbers of every hospital in the region, all the numbers of the staff in the car showroom you happen to be buying from this week, every neighbour in a one mile radius who might spot something happening?

Maybe you rely on everybody calling one of your family members and they'll pass the message one, as you'll take their call; only, you told them never to pick up an unknown number...

Normal voices already baffle most speech recognition systems

How come all these bad guys get to use working speech recognition systems that can easily run a scam without the human always catching on but the banks and insurance companies are stuck with machines that can't tell the difference between "Yes" and "No"?

After spending one morning this week getting the 'phone banking to just put a replacement 2FA device in the post and another morning getting the insurance changed so we can drive away the new car, both of those thankfully had successful endings - when we managed to get the machines to give up and pass the call onto a human.

SpaceX's Starship explodes again ... while still on the ground

Re: I'm starting to wonder

> She's a m...

Shh. Don't say that word, you'll only set him off.

Anomaly

I experience an anomaly.

You have an incident.

He/She/It has a catastrophe.

We are all doomed, doomed I say.

You (plural) face an existential crisis.

They are smoke on the stellar wind.

American coders are most likely to use AI

Some niggling details missing from that paper

First[1], they describe creating a corpus of known AI generated code by using two LLMs in sequence, one generating an English summary of a function coded in Python, which the second takes and generates a new Python interpretation. This AI gen'ed corpus and a known-human gen'ed one are used to train their "Is this AI or human code?" classifier used on the Github repos.

BUT they do not say whether or not any of their AI gen'ed code works, or is even syntactically correct - whilst you would assume (!) that commits to a repo would at least do that. By not mentioning that at all, they also miss whether AI gen'ed code needed to be patched by humans before it was commit-worthy and whether such patching changes the pattern in the code. So, is their classifier actually trained on the correct thing? Annoyingly, if their AI gen'ed corpus was duff then the classifier would probably be overestimating the amount of human code.

Second, they do not describe attempting to give the commits they found any kind of weighting when assessing the value - they are just taking every one of the 80 million commits and assigning each the same monetary value. So one AI using dweeb who continually generates trivial changes like rounding the corners of all the dialogue box controls will massively outweigh the human who made one commit in six months, but that change increased throughput by three orders of magnitude and/or corrected the calculation error that had plagued the project for years. But the latter project only had seven downloaders, whilst the dweeb's changes were showing on the screens of 15,000 other developers. Although those six downloaders were using the code across the websites of the four largest shopping websites in the world. In other words, how the bleep can they come up with any even vaguely realistic numbers for the financial value of those changes? They didn't even weight them by numbers of Github stars or monthly downloads - so were all those AI commits being made into repos for "my coursework on using AI" (or "the agency said AI must appear on my CV so I've forked 99 projects into my Github account, told AI to tweak them all - shame none of my pull requests have been acepted, but nobody will notice").

Third, they claim that the increased use of libraries by the AI gen'ed code means it is being innovative and therefore even more valuable. OR is it pulling in yet another JSON parser (that is the third one this week!) to call a web API to get the current date, bloating the program and causing even more network traffic? Then again, all too many humans tend to do that (oh goodie, you've added a new OOP wrapper ove SQLite, how novel) so maybe we're on par with the AI here, who knows? Well, not the authors of this paper...).

[1] no, not "first of all" as I am going to go on to mention another niggle.

Glazed and confused: Hole lotta highly sensitive data nicked from Krispy Kreme

US military ID numbers

Glad to know that they did check people were properly inducted into the military and had, presumably, passed basic training in the safe handling of dangerous equipment.

Some of the Krispy Kreme displays in motorway services make Dwarf bread look as harmless and innocent as a Mountstevens' Congo Bun. Stack up a dozen KK Signature Glazed Originals from that stand and it'll turn away a nine-inch Stiletto Croissant from Bloodaxe's Artisanal Bakery, 12, Treaclemine Street (now carrying a full line of waffle holsters and eclair scabbards; Bloodaxe Buns, You Can't Bite 'Em).

Firefox is dead to me – and I'm not the only one who is fed up

Re: OK, but what now?

> Pi Hole is pushed as the be all and end all

said in response to a comment that started

>> PiHole does ok at blocking ads, but...

So, hardly being pushed as "the be all and end all" - in fact, not "pushed" at all, most of the comments just say "I use it, get this result", and we're not even getting a voice-over by Stephen Fry, so not much of an ad budget there.

> various posters, saying very very similar things

Pi Hole really does one thing, does it the same way for everyone and works; it isn't some giant do-it-all-and-bake-a-cake-as-well bloated monstrosity, so everyone's experience of using it is going to be the same. Which is a Good Thing, btw. So, duh, everyone is going to say the same thing about it: it works, it is useful, you'll be better off if you get one, still use the other mechanisms as wel to forestall as much bad behaviour as you can.

> makes me wonder if there is a degree of sockpuppeting happening online...

I have better uses for my socks.

European consumers are mostly saying 'non' to trading in their old phones

Re: Also

Ah, you've been looking at stuff on eBay as well...

UK students flock to AI to help them cheat

> I could see using AI to write the actual essay text

That *could* be a reasonable argument, if the "AI" you used was actually a model of language (e.g. English) with a general vocabulary - i.e. the results of Natural Language research. But that was it - no big database/knowledgebase, no means of looking information.

> if you've done the initial research - organise your chosen stuff into bullet-pointed chapters / paragraphs (depending on desired length) and then have the AI do the actual language.

Such a program could then take the information you provide, and no other information, and turn it into an essay. Or a thesis. Which might be reasonably handed in as (almost all) your own work.

BUT you don't have access to anything like that theoretical program - what you can get hold of is an "LLM" which almost exactly the inverse: it has ingested loads of data and can spit it out again. Any language processing it has isn't the result of NL bods carefully declaring the grammar of English but was derived from those data loads and is inextricably linked to it.

As things stand, there is no reason for anyone handed an essay output from an LLM to accept that it contains only the information you provided and didn't pop in a few additions of its own.

Re: Snap test!

"The factory steam whistle was shaped like a parrot's beak. My aunt, who I live with, has a parrot..."

Penn State boffins create silicon-free two-dimensional computer

A 2-dimensional computer

and it's not called ABBOT? Running an OS called EDWIN?

Although that might encourage unkind comparisons between the shape of many programers and the sublimely, if not mystical, perfect shape of A Sphere...