iiinova — where your digital self keeps evolving

Various Earth orbits are a constrained resource that should belong to humanity and be managed by humanity collectively. Private, for profit space operations should not significantly add to the risk of a travelling public who gained nothing and had no say in accepting this risk.

110

But hey, SpaceX wants to do thousands of launches per year so that Elon can own space-based internet and profit from selling launch services. Does this smell like corruption to anyone else?

000

The old rule limited such transits to 6,412 per year to keep the risk of a fatal air travel accident from space debris to less than 5% per 80 years. Which itself was a kind of bargain between military and NASA launch interests and public safety.

100

Drífa the Snynx @drifa.bsky.social · 05/10/2026

Just before the SpaceX IPO, the FAA scrapped a rule that limited the number of aircraft flights under the parts of a rocket launch trajectory where debris from an exploding rocket were less probable, but still more probable than you’d like. (You still can’t fly under the highest risk areas.)

100

Failing to do this creates the false impression that real outrages are mere BS talking points. There is no reliable way to get deep MAGAts back in the reality-based community, but getting 80% of their neighbors so horrified that they feel compelled to hide is the only way we avoid doing a Brazil.

000

Drífa the Snynx @drifa.bsky.social · 05/10/2026

The Magyar part of this is particularly worth listening to. Possibly the biggest thing Biden &co. €£¥%ed up was not dragging the corruption of Trump and his acolytes through a lot of very public trials that result in penury and imprisonment.

100

Reposted by Drífa the Snynx

Robert Evans (the Only Robert Evans) @iwriteok.bsky.social · 05/10/2026

this is not a problem you can fix with calm, reasoned debate. you can't even fix it by improving people's lives with policy. there is no clear, proven, non-violent solution to this issue. finding one is the most significant problem of our age.

642192336

Reposted by Drífa the Snynx

ceej @ceej.online · 04/10/2026

“how do we win back trump voters in 2026” we don’t. we create an environment where admitting to voting for trump is so socially embarrassing as to render them completely inert. you don’t “win back” trump voters. you demoralize them into staying home. you make it a dunce cap

150107852333

Drífa the Snynx @drifa.bsky.social · 04/10/2026

Even beyond the stupidity of citing a BS bot as a source, consider this: the training data it is pulling from to weave BS that mimics human writing necessarily includes a lot of rapist apologia.

000

Drífa the Snynx @drifa.bsky.social · 04/10/2026

I had to see this cursed headline, now you do, too.

Headline: “NJ’s lieutenant governor told PBS, AI said he didn’t commit sexual assault.”

100

No involvement from human factors experts. Not a lot of experimental assessments of how people used to do the job to build on. No serious evaluation of the performance of human + LLM tool. Just … turn ‘em loose before the next guy does. This is a recipe for disaster.

000

But that’s not what’s going on. What’s going on is that people who don’t really know what engineers do—because they’re not engineers or just because humans are rubbish at self-reflection—are throwing descriptions of human tasks at chatbots and turning them loose.

100

Even with the best advice from human factors folks, reality can surprise you. Proposed automation, whether based on language models or anything else, needs to be carefully studied to see what the costs and benefits really are before we use them in high-consequence development activity.

100

The same might be true for a requirements checker. Every extra automation is one more thing humans must learn and manage. If it’s not perfect, we have to check on it. Those checks can take time or distract people from other kinds of problems. We can be bad at checking. Skills can atrophy.

100

But also, the effects of automation are not always as we might predict. We’ve spent decades proposing automation for things like air traffic control. And when we do experiments with real controllers, sometimes a feature to help highlight some conflicts backfires: adding it harms total performance.

100

Requirements engineers, like programmers, safety analysts, and, well, everyone else, do jobs that are far more complex than anyone realizes. We’re realizing this in software as bots can sometimes generate useful code, but have in no way obviated software engineers.

100

It’s entirely possible, for example, that a specialized model could help flag potential ambiguities in requirements statements so that engineers can fix them before they cause a problem. But (a) this is not the same as doing a requirements engineer’s job and (b) it’s possible it might backfire.

100

I should note that there are places automation will likely help. Bespoke models could be a good addition to current processes, *if* we are careful about designing the automation so that it helps rather than hinders humans doing a job.

100

Wait, you can buy medical patients?!?

Sign advertising “42% off medical patients.”

The question is, will we insist on real remediation? Will we reward the people who pointed this out? Or will the people whose deliberate irresponsibility caused the disasters just skate on with their careers as has happened so many times in the past?

000

There will be hand-washing of the “well, we couldn’t have known” variety. It will be bullshit: of course people could have known. I know several colleagues ringing alarm bells. They are being ignored, and that is a choice.

100

But by that point so much technical debt will have accumulated in the form of untrustworthy development artefacts that the choice is either redo several years of engineering work by hand or just accept the risk of disaster. There will be enormous pressure to simply accept this risk.

100

The result of this is going to be disasters. Some will be big, kinetic losses. Lots more will happen after those losses, when it becomes clear that a lot of the engineering effort of the last couple of years is simply indefensible.

100

I see virtually nobody seriously thinking through how what they are doing might go wrong and produce disaster, and equally little consideration for just how likely novel technology is to produce unanticipated failures.

100

I see vanishingly little understanding of how the standards have produced outcomes as safe as they have. People don’t know what independence really means, nor why standards require it.

100

I see people doing things like feeding the text of standards for safety critical development into the bots, writing a statement that they should follow the standards, and expecting that the result will be good enough.

100

There has been a mad rush to automate every part of the development process for critical things like vehicles using the synthetic text extruders. Lots of this seems fueled by fear that the competition will do it first, creating a sort of intellectual race to the bottom.

100

I’m listening to a bunch of nonsense about using BS text generators in safety-critical applications. It’s intellectual ipecac.

Nothing cool will happen as a result of this. Tech grifters meeting with the grifter government is only going to result in yet more treasure flushed away for no public benefit.

020

There is no intelligence left in the federal government. Adding BS problems to the BS administration does not anything super make. All this particular order is going to do is make actual scientists in the federal government look stupid when they present their work.

000

… If we can’t commit to something like that, we will continue to have the vast coercive power of the state under corrupt control, and that is unacceptable. Clean it up, get rid of it, or suffer. Those are the only choices.

001

… The effectiveness of investigation and prosecution here is key. It has to be near certain that one would be caught and punished before any magnitude of punishment is a deterrent. …

100

… It means that most of the USA’s lawmakers, policy makers, executives, and their supporters should be surrendering all their possessions over past conduct. No gifting everything to a spouse, it’s all gone. Poverty. …

100

… It means eliminating PACs. It means financial disclosures. It means mandatory divestment from investments. It means investigations with teeth nasty enough that few in USA government survive. It means restitution—past corruption needs to be rectified. …

100

Robin LeBlanc, from work @robinfromwork.bsky.social · 29/09/2026

I don’t hold out a lot of hope for states with representative democracies, but to make any such a viable path to a just future absolutely requires drastic measures to address corruption. That means things like rules about disclosing and recording every meeting. …

100

Reposted by Drífa the Snynx

The film is written, directed, and co-stars Hitoshi Ozawa, a legendary Yakuza actor who is a villain in Yakuza 0, who got his fellow legends involved. This was his passion project inspired by one of his best friends, a trans woman, and he worked with the community for representative casting.

3272272796

Reposted by Drífa the Snynx

Duncan Hothersall @dhothersall.bsky.social · 28/09/2026

If attendance at a major women's conference was down 80% due to a boycott fighting against trans *inclusion*, it would be all over the Times, Telegraph, Mail etc. This last weekend, attendance at a major women's conference was down 80% due to a boycott fighting against trans *exclusion*. Crickets.

132502831

Drífa the Snynx @drifa.bsky.social · 27/09/2026

An alien researcher doing field studies in Ocean City, MD might conclude that human society is organized around a religious ritual called “mini golf.”

000

Reposted by Drífa the Snynx

Elizabeth Bear @matociquala.bsky.social · 26/09/2026

Happy Stanislav Yevgrafovich Petrov Saves The World day to all who celebrate.

630897

Drífa the Snynx @drifa.bsky.social · 26/09/2026

Just wait till your shows get all the ads for prepaid funeral services so you don’t bother your loved ones.

hyperlexuality @baphomet.male.manipulat.org · 25/09/2026

Reposted by Drífa the Snynx

[extremely atproto dev voice] why do people pretend to hate AI? is it for attention?

0629

Drífa the Snynx @drifa.bsky.social · 25/09/2026

It’s part of the USA’s epidemic of @$$holes in the Aaron James sense of a person who “allows himself to enjoy special advantages in social relations out of an entrenched sense of entitlement that immunizes him against the complaints of other people.” They don’t feel the need to hide anymore.

000

Drífa the Snynx @drifa.bsky.social · 25/09/2026

Clutching a flat white like some kind of prayer for tranquility

000

Reposted by Drífa the Snynx

Paris Marx @parismarx.com · 24/09/2026

Just when I think I can’t hate OpenAI and Sam Altman any more, new details come out about how ChatGPT helped a shooter plan one of the worst mass shootings in Canadian history in Tumbler Ridge, BC. OpenAI made an explicit decision not to report the user to authorities.

motherjones.com

ChatGPT helped the Tumbler Ridge school shooter focus on guns, tactics, and terror, our investigation reveals

OpenAI’s chatbot fed the shooter’s violent fantasies and planning up to the massacre.

1533451490

Drífa the Snynx @drifa.bsky.social · 24/09/2026

Don’t you just love it when yet another fascist sack of €¥£# uses furry as a euphemism for queer? These guys want rigid control and conformance. I say, don’t obey. Don’t conform. Be soft. Be yourself. Be weird. Have fun playing musical chairs if you want to. Be the happy alternative to them.