yes, yes, it's impressive
But everyone knows the SI Measure of whether it computes is the Doom measure.
Well...can it run it?
250 publicly visible posts • joined 9 Jul 2010
They walk into the bar,
Proclaim they can run it better and serve the beer at a fraction of the cost to the consumer.
Then the politicians find themselves being served warm dregs and their wallets have been nicked.
When Fujitsu/Crapita are approached,the politicians are buying the rounds!
"Ofqual is examining ways to ensure coursework remains authentic as AI-generated submissions become harder to distinguish from student work."
I've had the pleasure of being told a story involving my friends daughter and a university.
She was, putting it bluntly accused of using AI for her dissertation. Whether it was part use or all that wasn't relayed. But the Uni started shooting first, no informal conversation, no formal conversation, simply harsh letters saying effectively "you used AI, therefore we're going to fail you"
The basis of their argument was that after cross-referencing it with some AI checker, it was deemed to be a certain % possible it was written by AI.
All credit to the lady, she gave a prompt "f ...you" then fought like hell. Providing ALL her sources and all her contemporaneous notes that she compiled for her dissertation, effectively dumping her years worth of supporting material on their virtual desk.
She won the appeal, but left a real sour taste in the mouth.
She was also diligent enough to save all that material and have it ready.
Some students may not be so fortunate.
Point being, while there is one argument about preventing AI being used there is also the argument about effectively using methods to sense check AI accusations.
The results of that survey are quiet wild and somewhat suspect.
Sample surveyed 200+ (although no detail on how many responded!)
92% of LEADERS expressed confidence in production readiness of AI generated code
BUT 81% report increase in production issues linked to AI code.
Which produces many inferences, the first, that leaders have no idea what "production readiness" means or secondly their faith in AI is grossly misplaced.
Despite the 81% citing production issues....52% (of the sample) report significant increase in development due to AI coding yet amazingly 68% believe the AI has delivered value.
More production issues, more cost, shifting constraints = value?
Inference - we're doing a lot more....but it's causing us trouble.
45% describe spending as predictable, not completely indicative if the other 55% are just spaffing cash into the wind. I know there are figures on setting hard limits and automated controls to manage spend and reading the recent API fubars - we know EXACTLY how those controls are going to go!
Interesting 70% say maintaining the test suite is the bigger burden now.
This is constraints surely. Prior to AI, the weakest link in the chain was the MITM - meatbag in the middle - so code only flowed as fast as the resources allowed, with perhaps a reasonable level of assurance of working capability, meaning the test suite was never a constraint.
Flip the AI coding tap, and you somehow have to manage the deluge, and suddenly the test suite becomes the constraint...the answer? Automate the testing (I refer back to the 81% reported production issues currently!) - which leads to the 53% increase in testing, security and deployment costs.
Summary of survey, purchase our AI management platform to manage your AI delivery...the circlejerk is almost complete.
..."the advantage of using the better model will only increase."
and herein likes the quandry. The 'better' model is iterating as such a speed, the incentive to fund and be on board with a model is fraught with uncertainty.
1) how long will the model I lock into be "the best"
2) All the funding currently is going into frankly those who appear to be shouting the loudest. Then you have things like Openclaw and Deepseek just pop up out the blue and upend things somewhat.
These AI companies are in a race to the top to have one AI to rule them all. In their wake will be a slew of companies that jumped on the bandwagon, locked in to a model that wont give them the superiority they thought would actually come.
- On a side note.
For companies investing heavily in AI because it's what your customers are demanding - just for clarity - have you actually asked your customers? Because as a customer of things, I'm finding a signification correlation between the introduction of AI agents into customer workflows and my irritation and frustration levels.
Simply
Us AI to review the mailbox and identify duplicate bugs submitted by AI
Ai identifies first unique submission and gives a sticker.
Then AI updates AI with "bug now known" answer
All duplicates get filed to /trash
Ai then congratulates AI on a job efficiently done
AI then fixes buggy code and releases update
The cycle begins again.
Linus sits back with a Long Island Ice Tea.
/s
Had a similar experience sorting out a payment issue for my Sons Uni.
The payment handler is clearly a US company, the ai bot on the line was manipulated after enough attempts to put me through to somebody.
Now normally you get "if you'd rather us call you back press 1 or stay on the line and we'll get to you"....then perhaps a 2 min time gap then the message again.
Not this service.
No sooner did it end the message..it started the message again. Cycling perpetually
After 10mins of that balls I pressed 1...and was called back almost instantly.
You'll bitch and moan and in 10 years time you'll still be treading these boards (assuming fate doesnt intervene)
You like your tech news like your systems.
Strong, dependable, barely changing...
But remember each and everyone of you...at some point looked at your system and went...time for the upgrade.
I guess i'm in the camp of "cant be arsed shouting at clouds"...
I did use Brave , and to its credit, it did a fine job of blocking ads. On top of my pi-hole perhaps the best browser so far in terms of that capability.
However, some El Reg regs gave me a few choice details on why Brave ISNT the most ethical choice for a browser
https://www.spacebar.news/stop-using-brave-browser/
So..
I've not spoken to my mother for ~2 years now, after she disassociated herself from my Brother and I, on account of us trying to aid her get from under the spell of a scam.
She is, in her mind, in a relationship with a hollywood a-lister.
Why this person needs xbox gift cards is beyond me.
The last contact was when we were interrupting her buying a car to be handed to said a-listers assistant.
Police have been involved but owing she is in full control of her faculties (really really questionable) they are unable to do anything.
I think said bar needs an effective SI unit.
There are currently several bars being propped simultaneously across a multitude of private and governmental areas.
Presently compared to the bar at the whitehouse, no ones bar , anywhere, has any cause for concern when it comes to lower limits and just pointing out said fact makes ones own bar positively stratospheric.
Now if i were an engineer with combined age & years service = 70, I'd be tempted to take that parachute and make sure my number is dropped on the bosses desk.
Because i'm going freelance once I leave and wait for the inevitable call for assistance and agreement to my reasonably-higher-than-previous day or hourly rate, and do what i did for arguably more money.
Yes its finite, but with a cushty parachute still intact...life may be fiscally better for me
They say opposites attract.
My wife has - what I refer to as her ESPN (Mean Girls!) whereby just tech in her vicinity will do weird and unpredictable things.
I as the resident nerd of the household will bask her in my glorious aura and those problems seems to vanish, making a complete liar of her.
Like for Like?
Surely this is the backwards way of looking at things? Is my x thing, as powerful as y thing? Who cares?
The way to look at it is, "what is it I want to do?" - start with your requirements.
If person can do A-Z of required things on their 8 year old comp, then the consideration needs to be "how much better can I do A-Z on the cloud comp?"
If the answer is not much (OR in some cases A-E can be done super fast, F-M can't be done at all and N-Z are about the same.) then the comparison is accurate and thus determined very expensive.
We are consistently pushed through the "latest is better" with forced obsolescence and slick marketing, when in fact, its simply corpos just wanting to sell more *shit*
The bubble will appear self-sustaining because it's not yet had a real test.
In the dotcom bubble, this appears to be a series of interest rate rises in the US AND at the same time Japan entered a recession....the rest we know is history.
Now it's very interesting how the techbros are propping up the US Markets and Trump is pressuring the fed reserve to drop interest rates, or drop Jerome Powell, but currently Jerome Powell is staying put and stable on interest rates.
The current "*special* military floperation" being undertaken in Iran and its full ramifications are not yet fully known.
It would be quite the implosion if that kickstarts the AI sell off
I've been using Vivaldi for a bit, but recently started to try Brave..prior to that I was a longtime opera user.
Vivaldi is good, I just had a couple of frustrating issues:
- Typing a full url into the address bar would sometimes redirect that as an input to my preferred search engine. Seemed to be certain urls - even with .com tld's but only fix was to disable search on the address bar.
- the built in ad blocker was good, but full block broke some websites and partial allowed too much through.
With Opera and Braves default adblock mode they hit the right balance of blocking ads but maintaining functionality.
So far Brave is winning out.
Exactly this.
AI should be a data stream to produce an input, which is then combined with other inputs, such as user knowledge/experience to produce a final output.
Because as you say AI produces outstanding stuff to someone without the knowledge (or understanding) its being used as Output by default.
Quelle Surprise.
Using AI to write arguments for a court hearing makes them no better at traversing a legal system. Unless I'm a lawyer reviewing said articles there is a high probability that said articles will contain falsehoods and inaccuracies, completely jeopardising a case.
Why would AI coding be any different?
Did peform a teams meeting with copilot transcribing.
To be fair it does a decent job, but then has a sort of seizure at various points and spits out whatever gobble-de-gook it feels like.
You only spot that 3-4 months later when you're trying to work out what "Langostine the bitch-goblet potatoes" actually meant
I'd toyed with Linux many moons ago and the last laptop purchase was win11. Not a gamer, more a general use machine it did ok, but the proliferation of AI into everything just felt the final straw, so I ditched for Ubuntu Linux.
I feel many more would, if the barrier to switching wasn't so high for non-tech people.
It was *reasonably* straightforward for me, having misspent the summer of my life in and out of pc cases and software.
But I like the freedom it affords now
notwithstanding all of the above. Whilst America's enemies were numerous, it could be safe to say threats and actions were largely contained within borders.
Only now - this administration is disenfranchising it's allies - to the point whether even they are questioning whether "ally" is a valid term- AND if that wasn't bad enough....after ripping up an accord with Iran back in 2016 (which by all accounts Iran was actually sticking too)....America (or more pointedly Trump (alongside Israel)) have opted to kick a fucking hornets nest.
The tragedy here, is great IRE will have been stoked...and America won't feel THOSE consequences during a Trump administration, it may be 10 or 20 years down the line...
Two reasons.
1: MS has spent the past 30 / 40 years entrenching itself everywhere corporate. You have generations of people brought up using some form of MS ecosystem.
Love them or hate them - they have also spent a large amount of that time, making things "easy" ..and integrated. So generational familiarity, simplicity and all under one roof approach - plus they uses their gajillions to bring hardware manufacturers onside to ship their shit as standard.
2: You have two choices for competitor for MS.
- *Nix - built by nerds, for nerds and utilised by nerds. (as someone who has started switching to *nix, I'm definitely finding this out!). Far superior in perhaps every way to MS - except usability and simplicity.
- Google (yes built on *nix) but want all your data to belong to us....
Is it little progress...or is it simply little progress down one particular skill tree?
All things being equal you'd love to see significant developments in all branches of STEM, but resources (human, financial, scientific) are finite so effort tends to be put into areas that give the best result (or return, let's not be naiive!).
That said, you often find when attention shifts back to a lesser-attractive form - such as these rodent...they can take advantage of advancements in other areas (computing/battery power and size) to quickly advance
Given that you are actually able to disable AI involvement with a simple switch should be something of a benefit.
Puts the AI integration into the hands of the user.
I have no issue with "features" provided they can be simple turned off (ideally off as standard would be more preferable)