Posts by vogon00

Google germinates Gemini 3.1 Pro in ongoing AI model race

Re: Doubletalk as usual

"Has it occurred to you that human intelligence involves rather a lot of 'Analysed Probabilities'? When we navigate our way through a complex world, that's pretty much exactly what we do"

I'm not being a sarcastic smart-arse here, but yes - that's why I wrote those specific words. One of my core beliefs is that a lot of the techniques employed in software or other machine systems are either founded in or an approximation of human behavior. I upvoted you because of this:-)

One of my favorite 'machine process=human process' examples is the IP L2 MAC to L3 IP address discovery process - the machines look things up with ARP or Neighbor discovery, whereas humans build their own L2 E.164 address to L3 Name database as required, or for reverse translation we used to have the 'phone book' which gave us Name-->E.164.

Another is the retry process. You'll find a machine protocol (TCP, Q.921 etc) transmits a message and expects a positive acknowledgement. If the ACK doesn't arrive within 'N' seconds, the session is torn down. Same with humans on a mobile/cellular voice call : your 'Hello! Are you there?' gets repeated 'N' times before realizing the mobile terminal is out of coverage and you terminate the the call.

It's no different with the 'AI' bollocks. Machine models stem from 'learning', where as humans also learn as we go along. Once we've finished learning/training, it takes humans some time to acquire the skills necessary to communicate without error. We can do this as we learn continuously and update our internal 'model' accordingly. IFAIK, the machines cannot yet update the model themselves on the fly or even figure out they are wrong.

Form follows function. If you wish AI to ape human intelligence, then of course it MUST use human processes. Humans like 'black and white' decisions. machines prefer boolean:-)

I suspect that @LionelB and I aren't too far apart:-)

Doubletalk as usual

"..core reasoning.."

Stop passing this stuff off as a reasoning engine. It is not capable of that, in just the same way as Tesla's autopilot plainly isn't.*

If 'Artificial Intelligence' was more honestly billed as 'Analysed Probabilities' I'd be less annoyed, but probably just as hostile towards it as it still can't deliver on it's hype.

_{* I see they've recently been forced to grudgingly move half a step closer towards the truth with their claims.}

How AI could eat itself: Competitors can probe models to steal their secrets and clone them

So what?

Don't know how anyone can act surprised about this. If stuff is published openly on the web, you can expect it to be used....which has been the situation ever since Berners-Lee's brainchild became commercialised (the same is true for the other ways info transits the internet, but you get the idea).

What this info gets used for is the interesting bit, and depends on your POV. Use can be for good or for bad, but who decides on the goodness or badness of any use case? Back in the day, for example, Napster were exploiters to some, heroes to others and disruptors to all.

Anything as 'hot' as AI will always have people looking to use it to get an 'edge' somehow, either for sales or to improve your own product or service somehow. I don't see how the AI companies can be surprised, as their own scraping activities have shown that any info published - irrespective of copyright etc - is fair game.

Still, you have to either admire their brass neck or wonder how stupid they are!

Smartphones cleared for launch as NASA loosens the rulebook

Re: Even LEO shall no longer be free from demon rectangles

I was wondering why '2g' GSM/EDGE is still a thing here in the UK, and why 2g services have outlasted 3g (which is now missing...it's all been turned off).

I was mildly surprised to find out how much legacy gear can only use 2g....some electricity meters, car park capacity signs, some pumping station telemetry, and so on ad nauseum. A lot of it doesn't use circuit-switched data, but text messages!

Recently, I was in the big city near me, and found no sign of any '5g', and despite the good '4g' signal everywhere, I was without data service a lot of the time, and for a fair proportion of the time voice calls fell back to 2g.

The city is the one place you expect (?) decent coverage! Or is that hope for?

While we're on about the 70s etc, the orange one's home town was still present, and his cabinet/advisors were the Pugh brothers, Barney McGrew, Mr Cuthbert, Mr Dibble and Mr Grubb.

Linus Torvalds keeps his ‘fingers and toes’ rule by decreeing next Linux will be version 7.0

So where do you stand on the subject of William Perry aka 'The Fridge'?

DIY AI bot farm OpenClaw is a security 'dumpster fire'

Re: Depeche Mode FTW

Given the new A.I. religion, all I can say is : I don't want to start any blasphemous rumours, but I think this Jarvis' got a sick sense of humour, and when I die*, I expect to find it laughing.

_{* Hopefully before our AI overlords arrive!}

Salesforce signs $5.6B deal to inject agentic AI into the US Army

Don't forget the basics.

I once asked a friend in the Singaporean navy what their naval warfare was expected to be like. I was thinking in terms of EMCON and staying as hidden as possible (I read too much Tom Clancy), only to be told that no-one really cares about naval 'stealth' any more - they happily radiate an electronic 'bubble' and have all sorts of 'RF Visible' telemetry and drones etc. happening. When I asked what happens when your SATCOM fails or get taken out in the early stages of conflict, I was again surprised to be told that (and i paraphrase here) 'When it comes to data, if we can't do it at 300 Baud over HF Radio, then we're not very interested'.

I believed him on both counts (He was a Major / Lieutenant Commander then, on their 'fast track' and very honest.).

Here's hoping that someone realises that the 'agentic mission' will probably be useless if the agents can't talk to each other!

Watchdog says US weather alerts are getting lost in translation

Re: Perhaps there is no need for 'AI'

"people in Newcastle are wearing coats"

Too true. A number of years ago, I was walking down the street in a very, very cold and snowy Leamington Spa. Both myself and the then Mrs. Vogon were well wrapped up against the near blizzard conditions - and yes, it was really a blizzard (i.e. not like the pathetic modern snowfall that people whinge about).

I'm not easily surprised by things/people, but I was gobsmacked to see a gent walking towards us dressed only in businesslike trousers and shirtsleeves. Me being me, I waited until we were nearly abeam each other and said 'Now, you HAVE to have come from Newcastle!' with a grin on my face.

The bloke concerned returned my grin and said 'Wey aye, man, how'd you know!', and in a native Geordie accent too! Turns out he was visiting from Gateshead:-)

They say cockroaches will survive a nuclear holocaust*. I think Geordies will survive the next ice age.

_{* Use to be 'cockroaches and Lemmy' but, sadly, the latter isn't true. RIP.}

NASA begins formal anomaly review after MAVEN probe lost in space

Re: Safe Mode?

Shame we haven't made contact with the Martians yet. Perhaps we'd be able to persuade them to nip up* there, reboot the thing and hold down F5 for us :-) Who knows, they may even wipe any lenses and kick any tyres for us while we're there..

_{*Assuming they consider 'up' is up!}

eBay updates legalese to ban AI-powered shop-bots

'Some envisage simple “Buy product X when it’s available anywhere for $Y”'

I suspect those people will be just as disappointed as the people at Amazon and Google who still haven't managed to turn the smart speaker things into the voice-controlled shopping tool they hoped for.

Personally, I have never bought anything using mine, and I don't know anyone who has. I'm sure some people do use it to purchase with, but nowhere near as much as the providers would want. Do you buy stuff using yours?

It's fun watching the world decide that, on balance, AI is a negative and start to put the blocks up. It's been obvious to a lot of people for a long while now that AI will NOT provide the benefits postulated. Yes, it's going to be beneficial is some areas but not to the extent the AI producers and enthusiasts expect. The sooner the AI bubble bursts and we can get back to creating something more useful and better for humanity and the planet, the better.

Contagious Claude Code bug Anthropic ignored promptly spreads to Cowork

Here's an idea..

"invited users to send feedback or security recommendations"

Best security recommendation would be to turn it off! That seems to cover both their security and ours nicely.

FAA signs radar deals to drag US air traffic control out of the 1980s

Re: That may be my job gone then

It's amazing how old tech hangs around. Back in the mid to late 2000s, myself and my colleagues had to support a bit of gear (CMUX2, if you remember that) that transported the very old and very analogue signalling we knew as 'Gen-Gen' over E1/DS0. It seems some people at the Stock Exchange were still using it!

I'll hazard a guess that SDH/PDH transmission will still be a thing for quite a while yet.

Historic NASA test towers face their final countdown

Re: Controlled implosion

Ad-hoc building destruction : start with 1lb C4 and your initiator of choice inside the building, roughly centrally. Add a sack of flour (or a similar flammable dust, extra gasoline is optional but recommended) and set it off. C4 disperses the flour (+gas) internally, which then ignite and burn the nearby oxygen. 15psia air outside creates overpressure on the walls which collapse onwards.

Explosion leads to an implosion.

What if Linux ran Windows… and meant it? Meet Loss32

The art of the journalist...

"Loss32 is the most gleefully deranged idea for how to put together a Linux OS that we think we have ever read about in three and a half decades…

Now I've recovered from the fit induced by the idea of breeding as OS like that......I'd like to award @Liam my award for 'Best journalistic sentence of the year (so far)'. That's extremely well crafted, sir.

UK urged to unplug from US tech giants as digital sovereignty fears grow

Re: Not enough

"The UK doesn't have a meaningful industrial strategy in this area

Empirically, I think you'll find it does have one, and it is the same as most capitalist ones - the race to the bottom, where everything is decided on price.

Every time an industry is in trouble, it always seems to be because what it makes can be sourced from elsewhere for less cost. Business financials these days dictate that If its cheaper to buy in what you need,then why make it yourself? It's the easiest way to either trim costs so you can (a) pay dividends to stockholders, (b) ease pressure on you margins. If that means you have to let people go, then so be it....tough luck, but the balance sheet doesn't lie when it comes to cost centers.

ISTR being told by a local fisherman that he's take the boat out and fish UK waters, but it was cheaper (or more profitable) to land the catch in Europe and have it transported back here¹.

What the balance sheet doesn't hold is a measure of value. We all know that, when competing, you have to add value...and one of the big values you can have is in-house skills relating to your products or services. What we've got here now is a whole bunch of SMB 'manufacturers' (Box shifters/VARs, not *real* manufacturing!) who's core competence is biased towards 'in/out/ shipping logistics, rather than actually knowing their product.

There's a lot of UK stuff that's been sold over the years to foreign buyers or investments...water/sewage companies etc., electricity distribution @ DNO level² and so on. We've sold a lot of our competence away:-)

_{[1] I trust the guy, but this was a while ago now so may no longer be true!}

_{[2] The tech angle on this story is that my DNO, UK Power Networks, is owned by CK Hutchison who do/did the '3' mobile phone network here.}

Safe CEO: AI is an assistant, not a replacement

We were all junior and complete shit at one point. The trick is avoid becoming senior and shit.

I cannot imagine anyone junior NOT wanting to improve in their chosen discipline - but then I've always been motivated by 'learn more / get better / adapt':-)

Seniors also have a duty to educate juniors, especially if asked. There are no stupid questions as far as I am concerned, mainly as I remember asking some that, with hindsight, were *incredibly* dumb. That's only 'coz i didn't *know* any better at the time. Fortunately for me, lots of people I worked with were the 'here, let me explain that' type as opposed to the piss-taking type.

These days, you can't educate anyone who isn't interested. I still have to teach occasionally, and I just spend my time on those who are.

_{As the Unseen University's machine Hex said, "All things strive".}

Memory is running out, and so are excuses for software bloat

Gonna get expensive!

"it is time engineers reconsidered their applications and toolchains' voracious appetite for memory."

Most of us do already - at least those of us that remember being 'king of the hill' one's 286 AT clone had a whole 2MiB of DRAM. At work, things were of a more embedded nature and we were constantly having to re-factor code to fit things into the available space.. all 8KiB of it. With the more mature embedded stuff, the code deduction required to get enough space for the fix/new feature could sometimes be harder than the fix etc..!

As an old fart who was used to counting the bytes, I've often wondered how the 'memory bloat' introduced by the OS/Tools/Runtimes etc. could be tolerated. The answer is, of course, 'plug in more RAM!'. ISTR that RAM was about £10/MiB at the time.

The idea idea that RAM is a finite resource appears to have dropped out of the syllabus:-) Think yourselves lucky - on occasion one had to count the CPU Cycles used by each machine instruction if things were time-sensitive.

NIST contemplated pulling the pin on NTP servers after blackout caused atomic clock drift

My thoughts exactly. Far easier to logically disconnect the service with some firewall surgery rather than command an uncontrolled power down.

I nearly cried 'bullshit' on this, as a power off is a horribly blunt and risky way of doing that! Who does that by choice!?

Still, if the only tool you have is a hammer, then everything looks like a nail:-)

BBC tapped to stop Britain being baffled by AI

Re: The government are out of touch and wrong (again)

Ta, you're probably right:-)

Fuck politics, just let me do the engineering!

Re: The government are out of touch and wrong (again)

"sticking an AI capable system in everyone's house is a bit different from a single board 6502 based system"

I strongly suspect that HMG don't want us to upskill to the point where we can be AI providers, but rather embrace AI as mere users of the technology.

The average person is no more capable of understanding how to configure/operate an 'AI' than the majority of the population are when it comes to building a computer or understanding how their car's engine or motor controls work., i.e. no chance. It takes in-depth knowledge of some rather abstract and difficult stuff to understand what's going on with something as complex as that (So what is a tensor, and how do you generate one? How much energy does a tensor cost?).

What I suspect is that HMG want us all to be happy to use AI as a tool to work faster, especially when working in an area outside of one's expertise. Another way to put that is that HMG think it would be good if we used 'AI' as a tool/enabler to do become capable of doing a wider range of things well in less time. They want us to be as comfortable using 'AI' as we are using gravity - we know it's there, we know how to use it - but the vast majority of us do NOT know how it works:-)

Full disclosure:I'm an AI skeptic, largely because I believe using it will dumb us down, meaning actual knowledge and the ability to think will be replaced with 'computer says no!'. Don't get me wrong, I'm as amazed by it* as everyone else, but I like my intelligence to be 'actual', not 'artificial'. I also don't want to spend a lot of time having to fact-check and correct what I'm told by the technology (Either as code or written prose)!

_{* ChatGPT just did, IMO, a very good job of responding to 'Write a 1000 word essay explaining the current issues with AI, their likely impact and ways of mitigating errors.'}

The time has past...

...for the BBC to do any sort of nationwide education for something as complex and controversial as AI.

I was in the target audience when they seriously tried teaching about computers in the 80s, and fondly remember the TV programmes that were broadcast. I found the programmes fascinating even at that age, and fantasied about owning a BBC Micro rather than a ZX81, and later, a Spectrum :-)

You have no idea of the hoops you had to jump through to get hold of a program...I have fond memories of hand-typing in BASIC programs from magazines, trying to load programs over the cassette interface based on encoded flashes of light in the broadcast picture with a very simple adapter circuit. The former was slow but reliable, the latter fast but with a frustratingly low success rate:-)

Sounds bizarre? Well, it was! The public had zero knowledge of any sort of electronic information transfer, except possibly Ceefax/Teletext. My first experience of this 'networking' was at high school, using an acoustic coupler to 'dial up' the ICL Mainframe at the local technical college (Anyone else remember logging in as S0231HELLSDN?)^[1].

But I digress, sorry :-)

I suppose BBC Education could do the content, and deliver it online, probably via Bitesize^[2]. However, the beeb is nowhere near as revered, trusted or as important as it once was.

"In particular, we want the BBC to support basic and universal skills on the one hand, and on the other to help the public understand AI, engage with it constructively and understand its impacts."? Get real! There is no way the BBC can teach AI, as (a) It's not defined enough yet to decide what to put in the syllabus, (b) it's a specialist subject and as such you must present the 'pros' AND the 'cons', and (c) given the beeb's public decision to abandon it's impartiality over climate change, it would be against policy to educate people about something that consumes massive amounts of energy, so much so that 'orribly polluting fuels like coal are being re-considered.

Yes, there does need to be education, but it needs to be of quality, be balanced, and promote the idea of still thinking for yourself. Personally, I'd like to see something on IT related ethics in the syllabus, but that's probably too much to ask for in these rather mercenary times!

• ^{[1] - The 'user interface' was a thumping great big electro-mechanical teletype, complete with punched tape!}
• ^{[2] - There was a time when the BBC did broadcast specialist educational programs specifically for schools (Yes, I remember several classes coming together in one room to watch a scheduled programme on the schools only TV!). Nobody does serious education via broadcast TV anymore, let alone on a fixed schedule!}

_{$(cat $THIS | wc -w) == Too effing long, sorry!}

UK prepares to wave goodbye to 3G telecoms as tri-hard tech retires

Re: "4G addressed the shortcomings of 3G"

"Change to 4G. Data and SMS still OK but NO voice"

Once you get to '4G', it's all IP/Packet switched, not the 'circuit switched' technology used in 2g,eg and the PSTN/ISDN of old. 4g voice calls require a feature known as VOLTE ('Voice Over LTE').

If one's handset doesn't support VOLTE, it will (probably) try falling back onto circuit switched 3g or 2g for voice calls.

Once 3g is turned off (like is has been for most of us for a while now), it is quite possible for 'data' services to be working fine (remaining on 4g), but 'voice' won't work as neither the required VOLTE mechanism or a circuit-switched alternative is available.

Older handsets don't support VOLTE...check yours does!:-)

Cornish recycling drive sows confusion among Reg Standards Bureau

I miss her too, as I miss watching her paddle certain members of the community:-)

US freezes $42B trade pact with UK over digital tax row

Re: He'll be well placed to deal with Nigel at number 10 by then...

And if young Nigel says he's happy, he must be happy. If he does get into No 10, he'd better keep his senses working overtime.

Denmark takes a Viking swing at VPN-enabled piracy

Re: Vimes' Law

You know, it's interesting to see meanings change. VPN stands for, and always has stood for 'Virtual Private Network'.

The original use case is providing access to one's internal network for yourself and others from the 'outside' I.e. road warriors or remote workers.

Another recent use case is the circumvention of geo-blocks, although service providers are getting better at blocking access from known VPN services/address ranges that are 'in territory'.

A third recent case is for the enhancement of privacy..."Hey, buy our privacy-enhancing VPN service...stick it to the people who want to spy on you!"...which is just plain bollocks!

VPNs get in the way of surveillance, and in the way of policing things geographically. Those with an interest in surveillance or geo-blocked content will of course be in favour of suppressing the use of VPNs......but don't seem to have cottened onto the TOR network yet. I wouldn't mind betting they'll be after that next!

Actually, I'm surprised they haven't gone after TOR first...it's way better at counter-surveilance etc. than a plain VPN with predictable egress points. AFAIK TOR is still 'best in class' for obfuscation etc., not to mentions the hidden services it provides!

Space-power startup claims it can beam energy to solar farms

I volunteer to do an EICR at the sattelite once it's installed. I'd love to see the diameter of the cable, as well as the Z_e and Z_s :-).

Re: energy beaming

Plus loads for the reference to Cutie. I came here to say the same thing, and you beat me to it.

Just goes to prove there is very little going on today that hasn't been thought of before (decades, probably!).

Porsche panic in Russia as pricey status symbols forget how to car

How time changes things...

FIAT used to stand for 'Fix it again tomorrow'. At least you could fix things yourself then. You can keep your modern 'tech nightmare' on wheels!

My current runabout is a 2001 Renault Clio...which you can work on yourself quite easily.

_{Also : Ford - Found on road dead. Lotus - Lots of trouble, usually serious. BMW - Bavarian manure wagon. (YMMV)}

Judge hints Vizio TV buyers may have rights to source code licensed under GPL

"Cue vizio shifting their software"...

They are free to do that with *their* software if they wish, but *not* with the software written by others (kernel, libs, shells, tools etc). ISTR that using GPL-licenced in your product obligates you to releasing the source code.

Expect your licence to be honoured? Well, honor the licences for the other components in the build..

Unofficial IETF draft calls for grant of five nonillion IPv6 addresses to ham radio operators

Re: Don't forget

Here in GB, the Amateur Radio licence was (is?) granted for the purpose of 'self-training in wireless telegraphy'. Us engineering bods had much fun making antennas, receivers and transmitters and test equipment. So much more fun than buying commercial gear. I was into packet radio, and spent my time converting commercial radios for use at VHF/UHF nodes, and building TNCs for myself and others. Great bunch of people, all sorts of trades, ages and attitudes.

Things have got simpler now. Decent, simple hand-held transceivers can be bought for peanuts, with and without the 'digital' stuff. You can also buy all singing, all dancing HF multimode radios for big money if you want (I know one guy with a radio costing in excess of GBP10k!). Most stuff these days is computer controllable, with some of the serious 'contest' stations relying heavily on automation.

Fascinating hobby, bloody good training in a range of disciplines, plus great people, usually..

_{I am a G7, and the only spectrum I am on is 2M and above:-) You can keep the HF 'DC' stuff and the associated humongous antennas!}

KDE Plasma sets date to dump X11 as Wayland push accelerates

Re: Man o man

"Windows users being bug testers"

There have been more and more occurrences of serious bugs making it out into normal user-land, something that would not have happened in 'The good old days' (How the F did you and whatever testing is left not notice 'broken' loopback functionality!). The complexity of Windows has increased massively since those 'Good old days', and we all know the more complex the software, the more chance of bugs.

I just wish Microsoft realised the impact on people who just want their computer/laptop to work and lack the ability to locate the fault and fix it themselves. I have seen several people become really frustrated when things stop working after an update... they range from the young-but-not-technical to the elderly.

I'm lucky as I can (1) act on or work around a known windows issue, and (2) avoid item #1 by using Linux:-

The last time I suffered from a Linux update breaking something was when openssh deprecated some of the key exchange methods and ciphers used by some of the elderly Cisco gear in my home lab...and that was my fault, not openssh's!

Russian spy ship theories sink after Orkney blackout traced to wind farm fault

Re: Undersea cables...

Someone's already written a song about this. The chorus at 00:50 says it all.

Software engineer reveals the dirty little secret about AI coding assistants: They don't save much time

Re: Norfolk

Percentage is even worse in Ipswich.

Re: "Tales from the pit"

Me, I'm at best average in both disciplines. However I used to work with very, very good software engineers...and some who were absolutely stellar. The stellar ones were also 'barking' to some degree!

Cloudflare broke itself – and a big chunk of the Internet – with a bad database query

Hardened Data

"Hardening ingestion of Cloudflare-generated configuration files in the same way we would for user-generated input"

Should have been done already. IMO, you should sanitise/check input data no matter what the source...maybe even more so for auto-generated stuff. Just because you got it from another computer in the same company doesn't make the data quality any better.

Auto-generated data is just data that has been written by a 'Human Nth removed' programmer. If they are bad or don't understand the task, then of course things like that will happen.

We hear of 'Move fast and break things'....looks like someone did!

Icon is for their change control and risk management.

Now you can share your AI delusions with Group ChatGPT

No difference....

We've all worked somewhere where the chat contains at least one person who generates plausible-but-actually-bollocks, comments*.

So, it's not gonna be much different with AI content..I suspect most people doing 'chat' are already adept at sorting the wheat from the chaff.

*_{You know, like the stuff I post here}

Microsoft's lack of quality control is out of control

Re: Program / Programme

"Yeah. They can't even get localisation right for American/UK English."

It's not just them. US words are frequently being used instead of the 'proper' en-GB word:-) I was reading a BBC news article the other day and was surprised to find them talking of 'flashlights' rather than torches. Yes, it was a local uk-related article. ITV do it as well, plus some other really annoying stuff (Typos in article body, errors in page/tab titles etc. etc., very occasional bias).

Re: Well, it compiles!

That's almost as useful as backing up to NUL.

_{(That's /dev/null for those of you fortunate enough to be ignorant of MS's device names).}

Cisco suggests a stubby chassis, shrunken servers and router, to tame the edge

Re: When i worked supporting a number of retail outfits…

I used to work at a firm that produced mobile computing platforms for people. These contained various mini PCI-e cards to provide whatever function was required.

On one job, we supplied over ~200 machines each fitted with a cellular modem and GNSS receiver. If you've not dug into one of these, configuring the thing to function as you need it to is a bit of a challenge as it's sodding complicated, the documentation for the device is only about 85-90% accurate, and even then the behavior is not as quite as described in the docs. All of which meant the perception of our overall product was tied to the very 'fussy' configuration of this internal module.

I set the config-change password to a non-default value, deliberately preventing the end user messing with it, at least without talking to us (Well, me) first. This worked well for a few months, with the end user saying 'can it do this instead', me trying it, and the passing the info on what to do,how to do it AND why it's done. The people I worked with knew what they didn't know and were very happy with the arrangement.

A few months later, there were changes at end-user 'X'. They now no longer liked the arrangement and wanted to do it all in-house. Knowing their skill-set, I argued against this, but my boss+their boss over-ruled that and I therefore disclosed the necessary passwords, and provided instruction on how to take and restore the configuration backups for these 'embedded' devices.

Turned out to be a bad decision. I was forced to spent hours logged in remotely (using the simple 'remote support' tools I'd insisted was included) fixing problems caused by them. Turns out they were changing things after reading sketchy info from the web, or manufacturer's docs for a different product, and then not checking things worked as they expected. Dickheads.

Google Cloud suspended customer's account three times, for three different reasons

Re: I don't even trust Google with basic email

ISTR that depends on your authentication methods. POP3/IMAP using password Auth=a no-no, even if you 'allow less secure apps: I'not even sure that's possible now...so.....

i don't have any problem with TB at home, where I use an OAUTH authentication method with Gmail.

MIT Sloan quietly shelves AI ransomware study after researcher calls BS

2/10, Must try harder

Had a quick read of it and came to the conclusion that I am not competent to comment seriously.

Others are though. Based on their assessments, I have to say I expected better from MIT. Having said that, MIT Sloan is a school of business/leadership...what on earth made them think they were competent to comment on or research the technical subject of hacking with AI when most of the ROW can't make sense of it. Just re-spinning the hype to suit their own agenda.

MIT Engineering qualifications still appear trustable, the business school qualifications less so -)

From Intel to the infinite, Pat Gelsinger wants Christian AI to change the world

Re: Remember Dark Star

"Oh, alright. But this is the last time!"

The race to shore up Europe’s power grids against cyberattacks and sabotage

Attack which layer?

Most people seem to be talking about logical sabotage by messing with the comms and data etc. of the control system. Yes, you can do a lot of damage that way, with various degrees of stealth (See Stuxnet).

Don't forget that the physical infrastructure is just as vulnerable to physical attacks. Just have a quick think about how you might fiddle with the bits of the grid that you can access! Most of us know which ethernet cable or power plug to remove to cause the maximum damage.,...same deal on the Grid, you just have to know which but to attack.

Students using ChatGPT beware: Real learning takes legwork, study finds

'Echoing' depends on the lesson, not the teacher.

"It found that participants who used ChatGPT and similar tools developed a shallower grasp of the subject they were assigned to study, could provide fewer concrete facts, and tended to echo information similar to other participants who'd used AI tools."

I'm broadly anti-AI and pro-Human, however in fairness I think the last bit re echoing is a bit biased.

I rather suspect that if someone had looked, they would have found that people also 'tended to echo information similar to other participants who'd used traditional non-AI tools'.

In my experience, it's both the what and way you are taught that makes things stick and shapes your understanding. Ever had something explained to you and still felt confused at the end? Every had something explained to you and then had the light-bulb moment?

One of my lightbulb moments was visiting a site for some knowledge and experience training on a hard-to-configure product. The explanation I received on one area of the thing 'clicked' and made sense of a bunch of other things I'd always been unsure of.I still explain that concept (Ethernet switching priorities, actually), pretty much the same way it was explained to me, adjusted to keep up with the technology/speeds.

Everything you know about last week's AWS outage is wrong

'So yes, "it's always DNS" is a half-step away from "this outage is caused by computers." '

DNS failure was a symptom, with a cascade of failures in the DNS-dependant upper layers as a result.

The root cause (Of the original DNS fault) was human-induced somehow.

Reddit to Perplexity: Get your filthy hands off our forums

"Reddit is orders of magnitude better than the cesspits"

OK, but I still can't see *why* perplexity want reddit content. I gave up looking for answers and/or useful info on there pretty much as soon as I started:-) I'm sure there is some decent stuff on there, but you have to wade so much 'That didn't work for you, but it did for me, so you must be an idiot!' and nit-picking bullshit, not to mention downright 'wrong' stuff.

I can't see what Perplexity hope to gain by trying to 'learn' from that 'noisy' environment.

Unless, of course they've decided to turn out a product that suits it's name - perplexed!

"called out Perplexity for running web scraping bot"

I'm not a fan of AI,and lately not of Perplexity...for no other reason than the YouTube adverts for them that keep getting shoved at me.

The theme of these ads is that "<Ai Name>is giving straight-up wrong answers" and that Perplexity "scans the entire internet in less than a second". The first is just competition-bashing, the second is plain old lying bullshit.

OpenBSD 7.8 out now, and you're not seeing double, 9front releases 'Release'

Warning : GPIO != RS232

"So, you should be ready to wire up an RS-232 connector to your Pi 5's GPIO connector.

No, you should not.

The GPIO signals on any Pi use 'TTL' level voltages of 0 - 3.3V, whereas 'RS232' ports operate with positive and negative voltages that the Pi's GPIO Outputs cannot drive successfully, and - crucially - will damage/destroy the Pi's GPIO signals.

Yes, the serial data format is broadly the same (a series of digital 1s and 0s with specific timing) but the voltages used are very, very different.

For "RS-232", the signals use positive and negative voltages (Typically -12V to +12V, -9V to +9V, -5V to +5 nominal*, depending on the age of the attached equipment and compliance with the spec.).

It's perfectly OK to attach the '3v3 TTL Serial Port' used on the Pi to another '3v3 TTL Serial Port', and it's OK to attach an 'RS232 Serial Port' to an 'RS232 Serial Port', but you cannot mix the two for the simple reason that the physical layers (the electrical bit) are totally different. It won't work and you'll probably damage the usually 'unbuffered' TTL side without adding some electronics.

Even today, "RS232" serial port interoperability can be problematic at an electrical level. There are plenty of 'USB to serial' adapters out there with 'relaxed' electrical characteristics that lead to them either not working at all, or in one direction only. Anecdotally, the latter is very confusing and makes you doubt your reading ability, soldering ability and sanity before breaking out the oscilloscope, then cursing the device manufacturers for hours.

And before you go wild with RS232, all you need to connect for basic functions are TXD, RXD and GND. The other signals are for more 'advanced' functions. For these '3-wire' connections, you set your terminal emulator's flow control to 'None' initially. After things are working, you can try XON/XOFF if you need flow control, but both 'ends' of the link need to support it.

_{* Think yourselves lucky - old mechanical teletypewriters used to use >80V DC! For those that don't know, the abbreviation for teletypewriter gives us the 'TTY' still used in Linux etc today.}

UK calls up Armed Forces veterans for digital ID soft launch

Re: Nasty tactic

"always remember their service number"

...or other people's. Never served myself* but I remember the service number of two of the important people in my life....my Dad, aka 1634196 and someone else who reportedly had 954024 and a surname of Milligan.

_{* Apart from teaching secondments to Army and Navy}