Posts by Jason Bloomberg 3353 publicly visible posts • joined 8 Mar 2008
"We're committed to providing the safest home possible for our customers data. That's why we've undergone ISO 27001:2022 certification, the most highly-regarded global standard for data security. Our auditors are UKAS accredited".
"Rest easy knowing your data is held behind a closed network, on servers based in the UK, that are not accessible to the public
"Keeping your data secure is the most important thing we do at Beacon".
What did Microsoft do to put this researcher on a life's quest?
It's hard to say and both sides have a very different view of who is in the wrong. But I am pretty sure Microsoft threatening to set the legal attack dogs on the researcher would likely have been the final straw that broke the camel's back.
It seems to me it's a classic case of "if you want to be shits about this - I'll show you what shit is". I have seen many cases of people, companies and corporations indulging in behaviour which elicits anger and desires for revenge and retaliation, failing to recognise that potential outcome, or not caring, arrogantly thinking "that will shut them up, that will be the end of it".
Unfortunately Microsoft appears to have picked on the wrong guy, upset someone who is now determined to have their retaliation, their revenge, and is is clearly able to do so.
As with most conflict, it matters less about who started it, who escalated it, but how it is resolved. I can't see it will be unless Microsoft are willing to accept some responsibility for triggering this saga, even if they don't believe they did or should.
On the plus side we can be grateful the researcher is enacting his revenge in a manner other than physical violence or physical destruction. And we know this conflict can likely be resolved. If it were AI attacking in a similar manner that is not guaranteed.
I suppose it is possible to hide a watermark in plain sight when AI users generally don't pay much attention to the dodgy bullshit your average AI emits.
T H I S W A S GENERA T E D B Y AI
I have noticed there has been an increase in the use of hidden, non-printable and non-visible, characters which get dragged in via copy-and-paste these days. My spell checkers usually highlight 'unknown words', which aren't unknown, following those hidden markers which is how I have observed it happening.
You know I'm gonna cheer the AI on this one. Bug finding
Absolutely agree and also that people need to find and fix some bugs before others exploit them. The ongoing heist of Bitcoin from Coinkite's Coldcard hardware wallets has reportedly reached $130 million courtesy of a trivial but impactful fuck-up with firmware builds which has been there for years. AI is suspected of having identified the flaw - though Coinkite may simply be trying to cover their own backsides.
The problem seems to me that most don't have a good strategy for dealing with the tsunami of AI submissions, legitimate, useful, and not. I hate to have to say it but AI may be the solution there.
Maybe it's different in Japan and Asia but I can't imagine any capitalist bastards in the west choosing to make things fairer for customers than more profitable for themselves. Paying for the infrastructure to deliver that.
Maybe one shareholder might vote for that but I wouldn't bet on it.
All the car parks near me added ANPR so they could maximise revenue, not reduce it, or make charging fairer.
The only way to stop ransomware is to cut off the money
That's not entirely true. We could have an army of those with a "particular set of skills" who deal with ransomware scum, their pals and their enablers. They can keep the money if it can't be recovered but they won't get to spend or enjoy it.
The only legal obligation on victims would be to phone "the number" and provide all the information asked for.
Then leave victims to decide for themselves how best to deal with the situation they find themselves in.
I'd also expect it to be made known that anyone who doesn't admit to being victim of a ransomware attack, or failed to protect against those attacks as best they could have, may also get a visit. But we can thrash out the details once we have decided to get serious about dealing with the problem.
I was offered the chance to relocate to some delightful backwater which was appealing given lower housing and living costs, even if networking would prove more costly and deliver less.
But I worried I would be stranded there should things take a downturn, my finances got squeezed, so I declined. A wise decision as it turned out.
even the mainstream press have started calling it out.
Which is a shame because I preferred their "it's so bloody dangerous it needs to be nuked from space" perspective.
It also misleads their readers into believing shit won't happen unless they let it happen - When few have the skills or means of preventing shit happening.
We need more "nuke it from space" headlines, not fewer.
He's doing it to pump up SpaceX's stock price because a lot of suckers believe this insanity.
I am still not entirely convinced this isn't the greatest piece of Performance Art ever, that Musk isn't doing all this just to show the world how stupid some people are. So stupid they will believe anything and even invest money and bet their futures on it.
"Then I did those Nazi salutes, dismantled America's institutions, encouraged people to vote for fascism, promised so many impossible things I thought everyone would catch on. But no one did. I wielded a chainsaw, got addicted to ketamine, started masturbating into test tubes, fathering surrogate children. The joke names for the kids were meant to be a hint, but nope. I kept promising FSD which was never going to happen, built the Cybertruck, promised humanoid robots, neural-link, and so may other things, then invented Grok. And then I told them I'd put data centers in space, said Space X would set up shop on the moon and generate revenues a thousand times larger than the current world's economy. I lied about everything but the suckers still believed me, kept handing me money. It was hilarious".
Maybe that pushed things over the 'fuck that for a game of soldiers' tipping point?
Unfortunately open-book completion with increased difficulty changed more than one parameter so the actual reasoning for the resultant AI slop is hard to determine, makes any conclusions drawn merely speculative.
Turning in AI slop doesn't equate to diminished cognitive ability. It just means that cognitive ability wasn't revealed by what was handed in.
The decreased final results may be a direct result of increased difficulty leading to students not fully engaging in those mid-terms.
In day-to-day operations, the UK’s Trident programme is completely independent from the US. The British Prime Minister can launch the missiles without any external input."
That may be true. But how far independence goes will have to factor in the US threatening to use theirs against us if we choose to use ours - That is no longer the ludicrous nonsense it may have seemed in times gone by.
And, while that talks about "day-to-day operations", I believe actually using nukes would fall far outside that scope. The PM may be able to launch them but that doesn't mean they will detonate over the intended target.
To me it sounds like a disingenuous statement to reassure Brits they haven't spunked our money down the drain, to pretend we have more independence than we actually have - Independent right up to that point where we no longer have that independence.
No one is going to admit we handed billions to a foreign nation for a nuclear defence which won't be usable when it needs to be used.
I have always wondered what payment would be required for a one-way adventurer or Darwin Award candidate to step forward.
I believe it would be a lot less than most would think.
The **really** tricky part is getting the proposal beyond the ethics committee. The Trump administration would have a huge advantage there.
"Your account has been compromised"
"Miscreants have used $1K in credits you are liable for"
"We have increased your limit to $10K"
"Miscreants have used $10K in credits you are liable for"
"We have increased your limit to $100K"
Given they could have blocked or limited the account as soon as they noticed something was amiss it appears Google are actively facilitating criminality, have chosen to be part of a criminal enterprise.
Who could ever have predicted that?
My understanding of on-prem "blockchain" means logging transactions to provide an audit trail while applying cryptography to prove its integrity and prevent tampering.
Being a more secure version of transaction logging with incremental checksumming I would have thought there would be little which could go wrong.
But I have no idea what ASX meant by "blockchain", how they were using it, or how they managed to screw up so badly.
I am sure they'll have better success going forward with "AI" now this fiasco is behind them :)
I believe a lot of it comes down to flawed metrics, trying to measure things thought to show staff are doing their jobs and aren't taking the piss - When taking the piss really doesn't matter so long as they are doing their jobs and getting things done on time.
I have found the more you let staff choose for themselves how they do their jobs the greater the rewards in terms of 'best boss ever' and in willingness to go the extra mile when that is needed.
One of the reasons that is not popular is probably because it means those managing staff have to be good at managing, can't just point to a full office and say everything is hunky dory.
There'll be a long list of projects that were delivered on time and in budget because those aren't the kind that make the news.
That's often said and it may be true, but it's hard to find even a short list, and I admit I would struggle to name successful projects.
Given the usefulness of an "it isn't all bad" list I am surprised nobody has produced one. Maybe it is because failures are so extreme that successes won't make a dent in those.
I would love to see a list of projects, how much they were expected to cost and how long they would take, alongside actual cost and time taken, how much they were expected to save and how much they did save or cost.
A column indicating if they were supported by the public, or predicted to be an absolute disaster, might show some interesting correlations.
How many major projects achieved what they were meant to? How many failed, as had been predicted?
The "RPD" release has two things going for it - It is really easy to install on older PCs, laptops and netbooks, especially 32-bit, and is better than good enough.
While it's an old "Bullseye" release I don't see "Bookworm" or "Trixie" adding anything essential for me (YMMV). The biggest problem I foresee is when it stops getting security updates.
The modern variant of Re-Education Camps.
But I don't see there will be many alternative jobs as AI increasingly removes workers. Even 'toilet cleaner' is going to take a hit. And the CxOs don't actually need that many punkahwallahs, arse-lickers and concubines, and certainly won't want any from the disgruntled pool, quietly bearing a grudge.
Construction, trade and blue collar workers, may be gloating while imagining their jobs are immune from AI but that will quickly change once the mega-corps have built the robots to do their jobs.
There will be minimal human involvement in building euthanasia booths.
The only jobs on offer will be as mercenaries for owners of volcano lairs. No salary but at least you get fed. And pension schemes won't matter much when fighting the killer robots.
Our only hope is to end it before it starts.
I've put off any. IT purchases for one year. If I need anything, it will be bought secondhand.
I expect that's the plan for a lot of people.
However, prices on the secondhand market are also rising. Companies and individuals delaying upgrades and not adding to what's available only helps that. If RAM prices remain high for two or more years, my bet is the market will be severely depleted and prices will reflect that.
Most consumers will likely settle for putting up with what they already have. Some might reflect on why they are so addicted that they can't just go without.
I fully accept the idea of not reinventing the wheel but the zeal with which libraries got dragged in to code has always concerned me. Especially when the coding avoided was trivial and those libraries often brought a fuckton of library dependencies with them.
But it was pointless to argue against "libraries, all the way down" bloat when memory was so cheap it was effectively free.
Mine's the hand-woven 1980's 'Told You So' T-shirt.
They are likely to be single board computers like the Raspberry Pi, where makers are choosing older boards in preference to paying the extortionate prices asked for top of the range, with manufacturers increasing their production of those boards. Makers will often choose more memory than they need when it is just a few dollars extra, as it's always better to have more than not enough. That ship has long sailed.
Embedded manufacturers of consumer products, who design their own boards, are probably also opting for older memory to keep their costs down, their sales and profits up.
There are consumers who will have bought older PCs and laptops and have set about expanding memory.
AI couldn't have taken off at a worse time with most of the world facing cost of living crises.
I wouldn't trust Burnham any more than any closet or overt Blairite in the Labour party. He has flip-flopped just as all career politicians do, and I find his desperate desire to be t'king o' t'country rather worrisome. I fear we will simply be moving from Starmer to Starmer-Plus-Plus.
Meet the new boss, same as the old boss, but with added charisma - https://www.thecanary.co/uk/analysis/2026/06/19/burnham-same
Most computing uses simple multiplication like you learned at school, with optimisation to reduce the number of steps.
I used 'quarter square multiplication' which made multiplication fast enough for my needs: ab = ((a+b)2)/4 - ((a-b)2)/4
Both halves can be lookup tables so best suited to small bit-size integers: ab = lookup[a+b] - lookup[a-b]
Presumably there's been a fair bit of heavy lobbying from certain parties to discourage this being made law.
I would hope so. Why should you or anyone be forced to absorb running costs or disclose IP still used in other products just because one product has been discontinued, is no longer supported?
That applies to any product or offering. While I sympathise with gamers it was just wishful thinking they would ever get their desires fulfilled.
The only way I can see it working is if businesses are paid to keep games playable after end of life and I can't see that ever becoming a thing. If you want me to keep an end of life product running you will need to pay for everything needed; servers, staff, office space, energy costs, insurance, and more. And who is going to pay those costs - everyone through their taxes, maybe to benefit just one user?
No, as nice as it would be, it's not going to happen.
If these tools genuinely made people more productive, you wouldn't need an "ambassador of innovation" to "galvanise energy and enthusiasm" and instil an "AI-first culture".
I presume the flawed thinking is that Civil Servants aren't using AI because they don't really understand how great and amazing it is, have maybe been put off by the Commentariate calling it shit, and they'd soon realise the error of their way if only they'd give it a try.
Good luck with that.
To be fair; A1 can ingest and wrangle unstructured content; it is simply that it costs a lot more for it to do that.
This all makes sense if a copyright holder wants to make it cheaper for A1 and users to leech off their content.
I would rather be going the opposite way, making it as expensive as possible, mangling A1-friendly structured documents so they are harder to understand. Just 'forgetting to include' some key XML closing tags might efficiently do that job.
This project is not meant to be used by Jason Bloomberg at all.
That means that Jason Bloomberg or any agent working on behalf of Jason Boomberg. What can be more clear?
"Is not meant to be used by Jason Bloomberg" doesn't actually mean I can't use it. "Is not to be used by Jason Bloomberg" would mean I can't use it. That would be more clear.
The two are not equivalent and the difference is crucial, especially if lawyers were arguing over that difference in court. If he intended to say "not to be used by" then why did he not just say that?
That "or any agent working on behalf of Jason Boomberg" seems to be a 'by implication' claim which I, or my lawyers, would argue isn't actually implicit.
This illustrates why I am supporting Plain English - Don't let ambiguity or wording which loosens the intent to creep in to contracts and licensing.
This project is not meant to be used by any "AI" coding agents at all.
Usage with any "AI" agent is strongly discouraged.
Neither are outright "must not" or "do not", and he also limits these clauses to "agents".
It is no surprise those dealing in AI shit have simply ignored the clauses and he may even have made himself culpable for the trap he laid and its consequences if one of those AI shits sets the legal attack dogs upon him.
I'd hope not but he's fucking with some very powerful and well resourced businesses.
Many portable and in-car DVD players had composite in and any second screen can usually be made to work with a mod of the cable.
It's not as glorious as a genuine 32" CRT but you can hold it closer to make it seem that it is :)
One thing I really liked about my Panasonic wide-screen CRT is that it stretched 4:3 more at the edges than centre so the 'short-fat people' effect wasn't so noticeable if you didn't want borders on the sides..
You had something like five years for that migration
Sure, but that did not change "it wasn't financial viable to move our free software to 64-bit for the few Mac clients there were".
If you are unable to use that time to migrate I don't see that as Apple's fault
Are you going to argue the same when it comes to Apple no longer booting Asahi Linux?
I don't care which way anyone argues it but it has to be consistent. Maybe we can simply blame the users for choosing Apple?
Does the computer owner's contract with Apple say that the company has to enable another OS to run?
I was going to say similar; do they have any obligation at all to support what users might like?
Apple screwed all the Mac using clients I had when they stopped supporting 32-bit because it wasn't financial viable to move our free software to 64-bit for the few Mac clients there were.
Oh well, how sad, never mind. That's the price of choosing hardware from those who believe in walled gardens. Sony ended support for Linux and other OSs on the PS3 citing "security concerns" and I would expect Apple can do the same using similar excuses.
I am not supporting Apple, or Sony, or anyone removing something others have come to rely upon. But if it's their right to do that there is not a lot one can do except appeal to conscience.
When you are all in the same boat, have to put up with it, it usually becomes normal. The western fad for washing hair and showering every day is a pretty modern invention.
I don't usually call anyone a snowflake but I do wonder how some would survive without modern personal hygiene luxuries, after the apocalypse, or even a week of wild camping.
I took the approach of not bathing or washing clothes during the pandemic to ensure my personal space was protected when I was out.
39,000 people arrived by small boat in the year ending March 2026,
And, when those arriving on small boats have their claims heard, the overwhelming majority have their claims for asylum accepted.
But you might as well be talking to a brick wall when trying to to engage with racists, or political parties who seek their votes, which shamefully includes Starmer's Labour Party.
A favorite ploy of scammers is to present something to you in an effort to part you from your money with a time component that doesn't allow you to consult your phone.
Any offer I see which is time limited, doesn't allow "I need to think about it", is a red flag to me, will see me saying "no thanks" by rote.
The only exception is when a trusted source points me to a 'special offer' which is ending soon which I might not have seen, and they don't care if I take it up or don't.
Remember back in 2024 when the bridge in Baltimore, Maryland was hit?
A couple of years earlier there was the Ever Given which went off-course and got herself wedged across the Suez Canal, closing it for a week.
The possibilities seem endless - Anyone up for an RC race through the Straits of Hormuz, a Speed 2: Cruise Control re-enactment, Death Race 2000 on the high seas?
