Posts by Moldskred

Mars astronauts could live in houses made of yeast and jello, say scientists

Re: Easier, faster, much more inexpensive answer.

Research _what_, exactly? While there are things to learn about Martian geology and, through that, planet formation and the history of the early solar system, there's just nothing we need a human presence on Mars to learn that can in any way defend the ridiculous costs and risks involved in a manned mission to Mars, much less a manned station on Mars.

A Mars colony is and will remain a pipedream. Not because it will necessarily remain technologically infeasible, but because there will not exist a sensible _purpose_ for one. Any human babitation it would make sense to build on Mars would make a lot more sense to just build on Earth, in orbit of Earth or on the Moon -- and so far we haven't even found much point in building human habitatio in orbit or on the Moon besides a couple of tiny space stations that are, really, mostly for show.

Agents made my retro tech safe to use again and showed their real value as testers of ideas

> 15 year old device with a tiny CPU, an ancient version of Arch Linux ,

A PC from 2011 might be old and underpowered, but "retro"? I mean, sure, you'll struggle with "Cyberpunk 2077," but it's all pretty much the same technology.

BOFH: Schrödinger's petty cash puts us in a very awkward superposition

Re: Quantum cash theory ...

Douglas Adams got there before you:

“Bistromathics itself is simply a revolutionary new way of understanding the behavior of numbers. Just as Einstein observed that space was not an absolute but depended on the observer’s movement in space, and that time was not an absolute, but depended on the observer’s movement in time, so it is now realized that numbers are not absolute, but depend on the observer’s movement in restaurants.” -- Douglas Adams, Life, the Universe and Everything

Waymo has designed a robocar chip to stay ahead of Tesla

Ahead of whom?

"Waymo has designed a robocar chip to stay ahead of Tesla"

Sure. Same as Chelsea signed Morgan Rogers to stay ahead of Sheffield Wednesday.

Tesla might be making a lot of noise, but they're so far behind as to not even be in the race.

Linus Torvalds says AI has made 'huge' Linux kernel updates the new normal

I'd be curious to know to what degree these are fixes generated by LLMs versus human-written fixes to problems found by LLM-based code analyzers.

AI researchers let models off the leash – then watched as they tried to add malware to a FOSS project

I'm shocked, _shocked_ that this slinky I put on the top of the stairs and gave a slight nudge, made an unsanctioned descent all the way to the basement.

Prominent Haskell defector pilloried by anti-AI purists

Where's the pilloring I was promised

How does the headline "Prominent Haskell defector pilloried by anti-AI purists" in any relate to the article?

Zuck's AI ambitions put Meta on course to become America's next big cloud provider

If you build it, they won't come

There's no general, large-scale _demand_ for raw LLM compute, though, so who are they intending to sell to?

Met Police boss threatens to cut 700 frontline jobs after Palantir deal blocked

"In the written version of his report to the London Policing Board on June 11, Rowley said the MPS has to reduce its full-time equivalent (FTE) headcount by 1,150 in the current financial year to balance its budget. The UOA would have covered around 500 of these by reducing staff time spent on backroom work including intelligence reports, mobile device analysis, and data processing."

A system that had not even been purchased yet was supposed to save 500 man-years of work in the _current_ financial year? Riiiiight.

Met Police's Palantir deployment has its own officers watching their backs

Re: Seems a bit servere

"I never expected the boot to stamp on _my_face -- forever" sobs man wearing boots.

To quote an old guy in a top-hat:

"I have always thought that all men should be free; but if any should be slaves, it should be first those who desire it for themselves, and secondly, those who desire it for others. When I hear anyone arguing for slavery, I feel a strong impulse to see it tried on him personally."

Atlassian’s new data collection policy protects rich customers while AI eats the rest

"In-app data would be the content created by users within the Atlassian platform such as titles and content in Confluence pages, titles, descriptions, and comments in Jira work items, custom emoji names, custom Jira or Confluence status names, and custom workflow names. "

As this will unavoidably include personal data, it beggars belief that Atlassian's wholesale properiation and processing of this data will not violate the GDPR on a fundamental level.

ServiceNow allegedly says salesman 'overachieved' and is not entitled to comp

Then what are they for?

And here I thought the entire _point_ of commissions on sales _was_ to make salespeople overachieve.

'Uncle Larry’s biggest fan' cut by email in early morning Oracle layoff spree

"But I’m someone who carries my laptop with me all the time and, if there is a problem that I’m working on, I don’t care if it is the weekend or Christmas, I’m going to work on that problem until it’s solved. [SNIP] I’ve seen stories about toxic work culture, but I never experienced it.”

Hmm. How does that saying go... "We don’t know who discovered water, but we know it wasn’t a fish."

Datacenter batteries are selling years in advance, because AI, says Panasonic

There's a lot of datacenters _planned_, but preciously few of them appears likely to actually be built. Once it all comes crashing down, there's going to be a lot of bag-holders. I hope they've requested money up front for all these batteries.

Amazon insists AI coding isn't source of outages

"While security incidents involving misconfigured access controls can occur with any developer tool – AI-powered or not – we have not seen compelling evidence that incidents are more common with AI tools."

I guess a follow-up question to that statement would be if they'd seen such evidence but not found it compelling, or if they'd not seen any such evidence and, if so, if they had looked.

Royal Navy races to arm ships against drone threat

Re: You owe me a new keyboard!

My instinct is to answer "beige," but I believe light weatherwork grey is the preferred choice.

Oracle vows 'new era' for MySQL as users sharpen their forks

Did he use to work at Microsoft by any chance?

"He said Big Red also planned to "extend and enrich" the developer community "

"Enrich." Because "embrace" would be too on the nose?

AWS intruder achieved admin access in under 10 minutes thanks to AI assist, researchers say

Re: Why an unencrypted bucket?

"Thinking quickly, Dave opens the lock in under 10 minutes using only an AI lock-picking robot, a squirrel, and a key to the lock."

Anthropic writes 23,000-word 'constitution' for Claude, suggests it may have feelings

Re: A genuinely novel kind of entity in the world

Look, the AI companies and their spokespersons have been doing their utmost to conflate LLMs with the science fiction concept of AI as conscious, self-aware, reasoning machine-_persons_ with feelings, motivations and desires. Yes, we all know that's bullshit, but I think it's eminently fair to hold the AI companies feet to the fire and point out that _if_ LLMs were what they're pretending they are, then it would be immoral to use them the way they are.

A genuinely novel kind of entity in the world

Ah. So they believe their business is based on a genuinely novel kind of slavery, then? Yeah, no, that tracks.

Vibe coding will deliver a wonderful proliferation of personalized software

Why "will", not "is"?

So why aren't we seeing a deluge of new software and small software startups?

Google previews Code Wiki: Can you trust AI to document your repository?

Obviously not

Useful documentation is provides context to the subject being documented -- you can't extrapolate documentation from the code itself because the required context isn't _in_ the code.

AI pilots keep crashing, mostly because firms skip the prep, survey finds

"You're doing Agile, I mean AI, wrong."

Microsoft teases agents that become ‘independent users within the workforce’

They never seem to say anything about exactly what they intend these agents will actually _do_ of useful work. Do they think that attending meetings and sending e-mails is the end goal?

Don't take AI to Thanksgiving: Bots have hidden biases

Sussman attains enlightenment

I'm reminded of the following koan from the Jargon File:

In the days when Sussman was a novice, Minsky once came to him as he sat hacking at the PDP-6.

“What are you doing?”, asked Minsky.

“I am training a randomly wired neural net to play Tic-Tac-Toe” Sussman replied.

“Why is the net wired randomly?”, asked Minsky.

“I do not want it to have any preconceptions of how to play”, Sussman said.

Minsky then shut his eyes.

“Why do you close your eyes?”, Sussman asked his teacher.

“So that the room will be empty.”

At that moment, Sussman was enlightened.

McKinsey wonders how to sell AI apps with no measurable benefits

80% decline?

"However, McKinsey also claims that the cost of inferencing is dropping rapidly, [snip]. The cost of large language model (LLM) delivery has declined by more than 80 percent per year over the past two years, it says."

No? The cost of inference is basically down to the complexity of the models, the price of energy and the cost of the GPU compute, and none of those have really gone down noticeably over the last two years, so where's this "80 percent decline" coming from?

ServiceNow thinks you're doing AI fast and wrong

"Chief Innovation Officer Dave Wright"

Yes, that certainly sounds like a serious position at a serious company that does serious, actual work.

Stop runaway AI before it's too late, experts beg the UN

And while we're at it

Let's also regulate personal backpack-sized fusion reactors, anti-gravity, time travel, faster-than-light travel, positronic brains, sonic screwdrivers and printers that just work.

Atlassian drops $1B on company that helps measure dev productivity

Developer productivity is bikeshedding

In the IT industry at large, the losses and inefficencies incurred at the company and project level far outweighs those at the individual developer level. Failed IT projects rarely fail because people write code too slowly -- they fail because people write the _wrong_ code. Any IT company or manager that's actually _serious_ about improving their output would focus their attention on the decisions-making process, goal alignment, internal lines of communication and project management.

It doesn't matter how many widgets your workers make an hour or how many hours a day they make widgets for when you don't know neither what kind of widget they are making nor what kind of widget they're supposed to be making.

Fiverr cuts 30% of staff in pivot to being 'an AI-first company'

Lavender-scented

This smells like "AI-washing" of a round of layoffs that was going to happen for other reasons anyway.

Fork that: Three alternative kernels show devs don't need Linux

Splitters

"What if a whole group of disaffected kernel developers, chased out of the project, wanted to go their own way?"

Didn't Monty Python do a bit about that in 'Life of Brian'?

Everyone needs an AI phone. No, don't hang up, it's true

The thing about voice-control...

Is that it's slow. Outside of pure dictation (which is a rarer use case than one might think -- for most text input you'll want to edit the text as you go) and situations where it's impractical to use some form of physical interface, voice control just isn't the better option.

Microserfs ordered back to the office, given 10 days to appeal

They can't get Teams to work either, huh?

"Importantly, this update is not about reducing headcount," she wrote. "It’s about working together in a way that enables us to meet our customers’ needs."

Search-capable AI agents may cheat on benchmark tests

Search-time is one thing

but if the benchmark datasets are available online doesn't that mean that they models are also almost certainly trained on these datasets, which would also make the benchmarks non-representative?

Google's $250 AI agent can only help you book restaurant reservations

Why is it _always_ making restaurant bookings? How often do these people book restaurants? After two decades of tech companies trying (and failing) to come up with "personal assistant" software solutions for the consumer market you'd think someone would have come up with another use case to try to sell to people.

SpaceX prepares itself for a tenth Starship flight test

Re: Making life hard

In this particular case it might be more of a "sunk face fallacy" -- less about the effort already invested and more about the _ego_ already invested.

Linux is about to lose a feature – over a personality clash

Apologies, my bad. That wasn't intentional; I must have misread the name.

I think you made a compelling argument that there is a conflict of personality and that Oversteer has been on the receiving end of unfair criticism, but the leap from there being a conflict to that conflict being the reason that Oversteer's patch was not included in 6.18 and that the decision to not include it was a "large [...] mistake in the kernel development management process" doesn't really seem to be bridged.

Re: Eh, sorry, what?

It also seems a bit optimistic to think we can accurately and objectively measure "technical excellence."

I don't know much about kernel development or filesystems and do not have a bone in the fight, but I have to say, Proven's comment doesn't seem particularly worthwile. He simply posits that Overstreet is in the right, that it's a large mistake to omit the July changes from 6.18 and that this was done because of a clash of personalities, without any explanation or further support for them. If you're going to publish a comment that someone's decision is a mistake, at least argue your case!

No more 'Sanity Checks.' Inclusive language guide bans problematic tech terms

Re: Censorship is bad for you.

Meh. Requesting that people use professional language in a professional setting is not censorship, nor is adjustments to what's considered professional language at an organization. Yes, occasionally people take it a little too far and the result becomes silly, but it's still not effing censorship.

GitHub CEO: Future devs will not code, they will manage AI

And yet, all that amazing AI-driven productivity hasn't seem to have had much effect anywhere. I'll start worrying about my job if I start seeing companies push out software at a markedly faster pace than they currently are.

Microsoft researchers: To fend off AI, consider a job as a pile driver

""It is tempting to conclude that occupations that have high overlap with activities AI performs will be automated and thus experience job or wage loss, and that occupations with activities AI assists with will be augmented and raise wages," wrote authors Kiran Tomlinson, Sonia Jaffe, Will Wang, Scott Counts, and Siddharth Suri. "This would be a mistake, as our data do not include the downstream business impacts of new technology, which are very hard to predict and often counterintuitive.""

Or, in other words, "Researchers working for company selling AI finds impact of AI 'possibly not catastrophic.'"

(Well, I say "finds", but that part's not actually a finding, is it?)

AI don't know: Enterprises slow to pick up on Copilot+ PCs

"The reception of Copilot+ PCs by businesses has been mixed," Canalys Research Manager Kieren Jessop told The Register. "Deployments have mostly been pilot programs to specific personas rather than mass roll outs."

If that's "mixed", I wonder what "poor" looks like.

The AIpocalypse is here for websites as search referrals plunge

Never go full Æsop

"A Man and his Wife had the good fortune to possess a Goose which laid a Golden Egg every day. Lucky though they were, they soon began to think they were not getting rich fast enough, and, imagining the bird must be made of gold inside, they decided to kill it in order to secure the whole store of precious metal at once."

Google's unloved plan to fix web permissions gathers support

Timeō Danaōs

And I'm not sure this even counts as ferentēs. The main problem here is that we just don't trust Google. We know from history that they've not been interested in protection user's privacy, so it's hard to give credence to their stated motivations here.

Canva to job candidates: Thou shalt use AI during interviews

Feels a bit ironic to be tested on how skilled you are at using tools that are being sold as not needing any skill to use, doesn't it?

AI can't replace devs until it understands office politics

Re: This is what I keep saying

"The AI figures out how to solve the problem and meet the requirements."

That's not how LLMs work. Fundamentally they do not analyze the problem nor check the generated code against the requirements. Rather, given an input text they generate code that a statistically plausible expression of that input text. "Plausible" here does not mean "likely to be correct," but rather "likely to _appear_ correct."

LLMs _always_ hallucinate. Occasionally, those hallucinations happen to match reality.

AI skills shortage more than doubles for UK tech leaders

"The number of UK tech leaders reporting a dearth in AI skills has more than doubled in the last year, according to research."

They don't even see the irony of that, do they?