Forgot your password?
typodupeerror

Comment Re:In response... (Score 1) 41

Altman (apparently) lies ... gaslighting

Gaslighting refers to a specific form of psychological manipulation and emotional abuse. It involves a sustained pattern of behavior intended to make the target question their own perception, memory, or sanity. The term originates from the 1938 play Gas Light (and its 1944 film adaptation), in which a husband secretly dims the gas lights in their home and then denies any change, convincing his wife she is imagining things and losing her mind.

This expansion of the term is widely documented. Commentators, psychologists, and linguists note that "gaslighting" is now frequently applied to:
Ordinary lies, Disagreements, Insensitive remarks, Political rhetoric.
Even forgetfulness or differing opinions.

When the term arises inappropriately, a calm clarification (e.g., "That sounds like a lie or denial, but gaslighting specifically requires an attempt to make someone doubt their sanity") can reinforce proper understanding without accusation.

Comment Re:Hardly the worst FAA idea of the decade (Score 1) 80

Is it true or not that the FAA was turning away qualified applicants for political reasons, and now has a shortage?
It seems to be very clear that the FAA has not been inducting sufficient trainees to allow for attrition, and has a shortage of graduates, despite plenty of applicants.
It is also public information that the applicant pool looks like a golf club - lots of white males, and there was an open policy of favouring the recruitment of females and blacks, in order to create a "balance".
    The controversial claim (from conservatives?) is that this policy has contributed to a higher attrition rate in the training program. It seems plausible, but is it true?

AI tells me: "The shortfall stems from a combination of structural, external, and operational factors rather than a lack of applicant interest. The FAA has received approximately 200,000 applications in recent years, yet conversion to fully certified professional controllers remains low"

I'll take "things that never happened" for $100, Ken.

That's not really much of an argument, it it? The fact is that parts of the FAA took the view that aptitude testing was racist, and deliberately chose applicants with lower scores on the basis of skin colour. You can argue over the value of aptitude testing, but it certainly is a good predictor of attrition rates.

Comment Re:Hardly the worst FAA idea of the decade (Score -1) 80

It's hardly surprising that we've been having so many air traffic control errors recently... we MIGHT be seeing the unfortunate collision of two problems: [1] prioritizing diversity over competence and experience, and [2] the long-standing FAA problem of simply hiring, retaining, and deploying too few controllers (which can mean not enough experienced people to monitor and mentor any new DEI hires who actually might work out ok).

"Too few controllers" flows from the first problem. By drastically cutting the size and quality of their applicant pool, and being forced to lower their intake standards, the rates of dropout and failure increased. (better than the alternative of lowering the pass grade) They might have predicted this, and compensated with a larger intake, but ideology stopped them from acknowledging it would happen. Lower staffing levels, regardless of race or gender, leads to more accidents.

Comment Re:Well, they _are_ our closest relatives ... (Score 2) 49

And also a scale of warfare orders of magnitude beyond any other species.

That's just utter nonsense, showing narrow thinking or unawareness. Obviously if you define "warfare" as how modern man resolves conflict when bargaining fails, you won't see it in mice.

Most of us are barely affected now by war. I've changed travel plans a few times, but thats about it.
Meanwhile, warfare is omnipresent at the microscopic level. A constant struggle of life an death under our feet, that we rarely notice. Or even at the macro.
Few species have such low rates of violent death, and high rates of living to old age.

    Some people just have to complain, no matter how good they have it.

Comment Re:Well, they _are_ our closest relatives ... (Score 2) 49

That is a bit misanthropic. Humans have achieved a scale of cooperation orders of magnitude beyond any other species.

Though you don't have to got back too far for when we had extended family structures like the chimps, that were subject to long-running blood feuds.
It was common in hunter-gatherers, and I'm sure still happens in some rural or tribal areas outside of government control. We are repressed chimps, but maybe that's good.

Comment Re:What I find amusing is... (Score 0) 38

LLMs don't actually know their own capabilities.

Those observations are somewhat out of date. Modern (ie 2026) frontier models have a lot more "knowledge" than their weights.

e.g. When I asked Claude about its own memory, it used a "product self-knowledge skill" which includes looking at its own SKILL.md file.
I believe Qwen 3.5 has similar capability, but of course you need to have it configured.

Comment Re:Could someone post the frustration regex code? (Score 4, Informative) 38

Ask Claude? He says:

This came out of the accidental Claude Code source leak on March 31, 2026, when Anthropic accidentally shipped a source map in their npm package exposing ~512,000 lines of TypeScript source code.
The regex lives in a file called userPromptKeywords.ts and looks like this:
/\b(wtf|wth|ffs|omfg|shit(ty|tiest)?|dumbass|horrible|awful|
piss(ed|ing)? off|piece of (shit|crap|junk)|what the (fuck|hell)|
fucking? (broken|useless|terrible|awful|horrible)|fuck you|
screw (this|you)|so frustrating|this sucks|damn it)\b/

Alex Kim's blog
As for what it's for: according to researcher Alex Kim, who first documented it, the signal doesn't change the model's behavior or responses — it's a product health metric to track whether users are getting frustrated, and whether that rate goes up or down across releases.

Comment Re:Linux desktop with 16 Mb RAM (Score 1) 116

Linux desktop with 16 Mb RAM was possible in the 90s

No, 2MB was never enough for a Linux desktop. I had 8MB on my 386 and it was only just sufficient.

Yeah, Bytes vs bits. But who measures RAM in bits?
I remember too 8MB being the minimum, but upgraded to 12MB so it was possible to do something else while the kernel was compiling.
How did we get to the point where 8000MB is considered a bare minimum?

Comment Re:Paradise! (Score 1) 200

The majority of automated speeding tickets in all AU states for which i found data was in the "0-9km/h" over the limit category.
They state there is a 2-3km/h threshold to allow for measurement errors. (though the machines now are extremely accurate)
Anecdotal reports are consistent with this.

Whats your reasoning coward?

Comment Re: Oh Brave New World with such people in it (Score 0) 137

You’re not wrong. Remember when they kept saying Kamala would start a war?

They? Gloating is unseemly. Nobody though Trump would start a war.
Certainly back in 2016, a small consolation over his win was that a new war less likely than under Clinton, who was something of a Hawk.

But back in the old days, when the US was a democracy, they would have needed support from Congress to start a major war.

Slashdot Top Deals

The moon is a planet just like the Earth, only it is even deader.

Working...