Trending
Beltane's Avatar

Beltane

@anbhealtaine

Working in tech, helping make AI safer and more privacy protecting

364
Followers
910
Following
461
Posts
04.08.2023
Joined
Posts Following

Latest posts by Beltane @anbhealtaine

Preview
Exclusive: US orders diplomats to fight data sovereignty initiatives President Donald Trump's administration has ordered U.S. diplomats to lobby against attempts to regulate U.S. tech companies' handling of foreigners' data, saying in an internal diplomatic cable seen ...

The US wants to pull the whole world down to its low level (often non-existent) protections of privacy and AI safety

www.reuters.com/sustainabili...

02.03.2026 03:04 πŸ‘ 3 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Preview
City shuts down police license plate cameras after judge makes footage public After a court ruled license plate reader footage is public record, one city turned off its entire camera network as lawmakers debate limits nationwide.

How increased transparency can protect from mass surveillance

www.wltx.com/article/news...

02.03.2026 02:51 πŸ‘ 1 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0
Two baby goats snoozing

Two baby goats snoozing

Kid nap

Snooze time for the new baby goats nextdoor

01.03.2026 18:39 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

JTAG, wow. That's a blast from the past

I worked for Rod Tulloss at AT&T Bell Labs back in 1990 when he was one of the people leading the adoption of the JTAG standard

It's nice to hear that it's still useful

27.02.2026 05:36 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Other alternative terms I'd propose to add:

contextual weighting (not attention)
pattern extrapolation (not hallucination)
pattern matching (not understanding)
iterative LLM (not thinking)
LLM workflow (not AI agent)
supervised labeling (not teaching)
context buffer (not memory)

15.02.2026 06:50 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Interesting paper proposing how we can de-anthropomorphize discussions of probabilistic automation systems (not AI) by using terms like:

errors (not mistakes)
text input (not prompt)
output (not answer)
conversation simulator (not chatbot)
weighted (not neural) networks
reflects (not shows) bias

15.02.2026 06:20 πŸ‘ 6 πŸ” 4 πŸ’¬ 1 πŸ“Œ 0

It's depressing to consider that Gramsci's analysis from his time applies to ours

"the old world is dying, and the new world struggles to be born: now is the time of monsters"

But he also called for optimistic acts in the face of this reality

"Pessimism of the intellect, optimism of the will"

15.02.2026 05:20 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Unfortunately, I don't think this is technically possible

Bluesky is built on AT Proto which is designed as a distributed system with no one organization controlling the data

That means *any* organization can see the data

15.02.2026 02:47 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Preview
Taggart (@mttaggart@infosec.exchange) Attached: 1 image What's going on here? The matplotlib maintainer this story is about correctly notes that all the quotes from his post in the article are made up. **UPDATE: Link was pulled; see bel...

This is even more interesting than I thought

That's yet another layer of AI craziness

The Ars Technica article itself may also have been AI generated

The article was pulled, which is why the link is broken

infosec.exchange/@mttaggart/1...

14.02.2026 20:41 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Screenshot from my news feed

Screenshot from my news feed

My news feed contained this item

It was about the behavior of an AI agent

And the illustration is, I assume, AI generated

And the news feed item was itself AI generated

But the link to the story on Ars Technica is broken, yielding a 404 error

So, AI all the way down yielding…

Nothing

14.02.2026 20:29 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Screenshot of where my account is in this space

Screenshot of where my account is in this space

My placement is a little surprising

Off in intergalactic space near some galaxies that are not really how I think of myself

09.02.2026 03:33 πŸ‘ 3 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Very cool

Did you give equal weight to followers and following?

Giving more weight to following may give more of a map of how people think of themselves

while giving more weight to followers might give more of a map of how people are perceived by others

09.02.2026 03:29 πŸ‘ 6 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0

Elon Musk vs. the laws of thermodynamics

I wonder who will win?

03.02.2026 06:14 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Though if we see a lot of things like this (which are more likely to actually work on the real threat model), then we know that the public is really concerned

bsky.app/profile/auti...

03.02.2026 05:01 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Good point

The fact that this company seems to think there is a market for this is interesting in itself

It may indicate a general growth in awareness of the privacy threat of facial recognition

03.02.2026 04:59 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

These will only work on IR-based systems in personal security devices (phones, computers) and access control systems

But it will be largely ineffective against mass public surveillance, which generally uses optical-based systems

02.02.2026 23:59 πŸ‘ 7 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Happy Imbolc, the celebration of the day halfway between the winter solstice and the spring equinox

It marks the beginning of Spring in Ireland

The ancient Celts were a bit pedantic about their solar astronomy, and didn't take account of thermal inertia which actually delays seasons by a month

02.02.2026 00:58 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Wow, appreciating this thread requires understanding a very specific combination of very online pop culture and political references from both Korea and Ireland

01.02.2026 05:54 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Starlink user data trains Grok AI by default

If you use Starlink you might want to prevent that

You can change the option under

= > Profile > Privacy

31.01.2026 06:35 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Remember: LLMs and bots and AI-generated things should always be called β€˜it’. Regardless of the style or tone of voice, there is only an β€˜it’ there. Don’t humanize it.

31.01.2026 03:49 πŸ‘ 268 πŸ” 51 πŸ’¬ 7 πŸ“Œ 3

If you're interested in AI policy you might find my new AI Policy feed useful

It's a filtered feed of BlueSky post relevant to AI policy and governance around fundamental rights, democratic institutions, and the rule of law

Click through and add it to your list of feeds

bsky.app/profile/did:...

26.01.2026 01:56 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Depending on their language's relationship to sacred cultural practices and the soul wound of historical loss, different indigenous peoples have different views of outsiders learning it

I believe Ojibwe, Choctaw, and Cherokee (in addition to Nahuatl and Navajo) do welcome respectful learners

17.01.2026 04:13 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I think you just change velocity and reliability are the main things you want to measure

Then expect all devs to spend some fraction of their time using their best judgement and subjective skills to continuously improve the code base -- which hopefully will case velocity or reliability to improve

12.01.2026 02:57 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Thing Explainer: Complicated Stuff in Simple Words

Remind me of xkcd.com/thing-explai...

06.01.2026 05:35 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Preview
Life expectancy vs. health spending The period life expectancy at birth, in a given year. Health expenditure includes all financing schemes and covers all aspects of healthcare. This data is adjusted for inflation and differences in liv...

American exceptionalism: pay twice as much for health care and still die 5 years sooner

Life expectancy vs. health expenditure before, during, and after COVID for selected countries

ourworldindata.org/grapher/life...

04.01.2026 20:53 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I'm particularly interested as I'm working on safety protections of a project to add personalization to a knowledge-seeking AI system

Personalization is arguably a kind of persona drift; it makes the system more aligned with the end-user

So we would need to distinguish safe drift from unsafe drift

02.01.2026 17:41 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Thank you for sharing this! Very interesting (and worrying).

I understand you don't want to share proof-of-concepts of "Socratic/metaphysical prompts" that you used, but could you share whether you actually managed to exploit this vulnerability to get a system to violate its safety protections?

02.01.2026 00:47 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

The language-emitting party of an AI system is indeed a bullshitter, optimized to produce fluent text

But it can produce grounded output by carefully controlling its context

Missing from the article is that useful AI systems fill the context of the LLM with grounded text from deterministic tools

01.01.2026 07:01 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

There is No Antimemetics Division was (ironically) the most memorable book I've read recently

It's such a mindfuck

I loved it

01.01.2026 03:40 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
It called for measures to attract migrants especially younger ones who will want to have children here. β€œThe core demographic group that typically fuels migration flows – people aged in their 20s – is projected to decline and will be increasingly in demand across other ageing societies,” in coming years, it noted.

β€œStronger integration policies, long-term settlement opportunities and support for returning emigrants are vital. A social contract of migration is needed where the state invests in integration and communities, and migrants are supported to contribute and settle.

β€œThis can strengthen cohesion and public confidence in migration,” it said.

It said that if its recommendations are implemented across Government, the results will include a stabilised population age structure, increased workforce participation, fiscal sustainability and greater capacity to invest in services.

β€œThis eases infrastructure pressures, improves the quality of life, and strengthens integration and retention of citizens,” said the report.

It called for measures to attract migrants especially younger ones who will want to have children here. β€œThe core demographic group that typically fuels migration flows – people aged in their 20s – is projected to decline and will be increasingly in demand across other ageing societies,” in coming years, it noted. β€œStronger integration policies, long-term settlement opportunities and support for returning emigrants are vital. A social contract of migration is needed where the state invests in integration and communities, and migrants are supported to contribute and settle. β€œThis can strengthen cohesion and public confidence in migration,” it said. It said that if its recommendations are implemented across Government, the results will include a stabilised population age structure, increased workforce participation, fiscal sustainability and greater capacity to invest in services. β€œThis eases infrastructure pressures, improves the quality of life, and strengthens integration and retention of citizens,” said the report.

Meanwhile Ireland recognizes the key to sustainable long-term economic health is ENCOURAGING immigrants, and actually competing with other aging societies for them

(From an influential advisory council that includes employers, unions, and community organizations)

www.irishtimes.com/ireland/soci...

31.12.2025 04:07 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0