Joanna Bryson
@j2bryson.bsky.social
9K followers 620 following 5.5K posts

Professor of Ethics & Technology @ Hertie School, Berlin. Interests: Artificial & Natural Intelligence; Behavioural Ecology; Cooperation; Digital Governance. Social media policy: https://joanna-bryson.blogspot.com/2024/10/guidance-to-my-social-media.html .. more

Joanna Joy Bryson is professor at Hertie School in Berlin. She works on Artificial Intelligence, ethics and collaborative cognition. She has been a British citizen since 2007.

Source: Wikipedia
Computer science 31%
Neuroscience 25%
Posts Media Videos Starter Packs
Pinned
j2bryson.bsky.social
thread. correct numbers and data backing them come later, but this is the thrust.
j2bryson.bsky.social
One thing I learned last week was that if you don’t count the BS market capitalisation bubble of 7 “AI” firms, the US economy as measured by the s&p 500 didn’t go up 13% last year; it went down 7%.

The EU may be a larger economy already. We need to get our own O2 mask on, then start helping others.
garymarcus.bsky.social
Guess why there was no reporting of AI revenue?

j2bryson.bsky.social
I think that he could have been treated unfairly positively and unfairly negatively both at the same time, though actually the article goes on to make Turing still sound like a hero for many reasons. But doesn't one move over to allow more people on the winner's podium?

j2bryson.bsky.social
I can't get the first, colour link to download? But re the abstract of the arxiv one, this was the most recent time I managed to get t EoL: joanna-bryson.squarespace.com/s/cultural-r... (2010)
joanna-bryson.squarespace.com

Reposted by David Underdown

j2bryson.bsky.social
Please see my pinned post here for a lot of data and discussion about this enormous AI bubble.
alondra.bsky.social
"Since the wealthiest 10 per cent of the population own 85 per cent of US stocks, they enjoy the largest wealth effect when they go up... the latest data shows America’s consumer economy rests largely on ...[the spending of the] top 10 per cent" #AI #irrationalexuberance www.ft.com/content/6cc8...
gdp1985.bsky.social
Over the past few years, Beijing has been mirroring the U.S. economic-security toolkit. China is developing its own architecture to manage risk, preserve chokepoints, and respond in kind. It's also learning from the United States.
alondra.bsky.social
"Since the wealthiest 10 per cent of the population own 85 per cent of US stocks, they enjoy the largest wealth effect when they go up... the latest data shows America’s consumer economy rests largely on ...[the spending of the] top 10 per cent" #AI #irrationalexuberance www.ft.com/content/6cc8...

Reposted by Joanna Bryson

grahamformaine.bsky.social
We have armed secret police kidnapping people off the street based on the color of their skin.

When we win: we will haul them before a Senate committee. The masks will come off. There will be consequences.

j2bryson.bsky.social
Sorry, my Sundays are often more desperate than funny (I'm writing lectures for Monday!)

j2bryson.bsky.social
Logging would contradict GDPR

(via @georgweizsaecker.bsky.social )
georgweizsaecker.bsky.social
This analysis is exploratory
unenthusiast.com
In honour of spooky month, share a 4 word horror story that only someone in your profession would understand.

rm -rf ~/

j2bryson.bsky.social
As time has passed & knowledge got tested, I was impressed by how much I did learn just by talking to smart people who indeed proved to be good sources. I think society has always functioned that way. What amazes me is how many people seemingly have no memory &/or no interest, but maybe that works.

j2bryson.bsky.social
Maybe follow fewer, more reliable people (or other news sources?) Seriously, if the rate is the problem, that's adjustable.

Reposted by Joanna Bryson

georgweizsaecker.bsky.social
This analysis is exploratory
unenthusiast.com
In honour of spooky month, share a 4 word horror story that only someone in your profession would understand.

rm -rf ~/
hammancheez.bsky.social
"The chancellor approved it"

Reposted by Joanna Bryson

hannesullrich.com
Remarkable product crippling by #Google, forcing auto-synchronization, a mediocre and unbearable #AI product, on #YouTube with no way to disable. Way to turn a potential feature into a bug!

j2bryson.bsky.social
The thing about reading in alternative newspapers or hearing in pubs is that you don't know for sure who to believe, but the things came out decades later in the mainstream news e.g. torture in Southside Chicago police stations; Thatcher tactics against the coal industry & unions. Now we see video.

j2bryson.bsky.social
This is simultaneously hysterical & terrifying, that Trump & his people will & can perpetuate blatant falsehoods re Portland or anywhere they choose.

So much for the information age? Though I learned a lot of suppressed truths in the 1980s from the @chicagoreader.com & the 1990s in Scottish pubs.
thedailyshow.com
The following is REAL footage from Portland, 2025. Viewer discretion is advised.

j2bryson.bsky.social
Not that "my" assumption must be right, but I thought that it's so widely shared it must have been tested and established as true or false by now? @glupyan.bsky.social's 2012 paper clarifies that nativists can ALSO be right without this being wrong cc @hugospiers.bsky.social @markgbaxter.bsky.social

j2bryson.bsky.social
We've known people were better at discriminating culture-specific colour-term boundaries since at least 1990s. I assumed neural recruitment to this important social task, and teach it alongside sensory vs motor homunculi reps en.wikipedia.org/wiki/Cortica... Mystified why this isn't established yet?
Cortical homunculus - Wikipedia
en.wikipedia.org

j2bryson.bsky.social
DeepSeek is training up way faster & better on the bibtex task. Hard to be certain, but as a reasonably-expert user, I suspect that it's not that DS has better fundamental AI, but rather Google Gemini has more "guard rails" aimed at manipulating users into treating their agents anthropomorphically.

j2bryson.bsky.social
I bet the only times people don't blame things on AI it's when they genuinely did vibecode something & don't want to admit it.

cf @mattgraham.bsky.social for a more US Pol take on the CDC fiasco. Apparently many people have been fired & rehired twice since the inauguration, many others really gone.
paleofuture.bsky.social
This government is a well-oiled machine that’s definitely not suffering from internal power struggles hastily papered over as technical mistakes
The employees "were sent incorrect
notifications, which was fixed last night and this morning with a technical correction," a senior administration official said. "Any correction has already been remedied."

Reposted by Joanna Bryson

paleofuture.bsky.social
This government is a well-oiled machine that’s definitely not suffering from internal power struggles hastily papered over as technical mistakes
The employees "were sent incorrect
notifications, which was fixed last night and this morning with a technical correction," a senior administration official said. "Any correction has already been remedied."

j2bryson.bsky.social
At @gotocon.com Copenhagen I heard a few people say that programmers are FINALLY making it easy to do things that ought always to have been easy – just so they can program "agentic" AI to do those things for them.

cc @simon.fedi.simonwillison.net.ap.brid.gy
michae.lv
Do the Zendesk agents resolve 80% of customer service issues in the presence or absence of a proper FAQ and easy-to-use and powerful customer user interface?
techcrunch.com
This has been a big week for AI companies signing enterprise deals, with Zendesk unveiling new AI agents that are supposed to be able to resolve 80% of customer service issues, Anthropic and IBM announcing a strategic partnership, and Deloitte also announcing a deal with Anthropic.

Reposted by Joanna Bryson

michae.lv
Do the Zendesk agents resolve 80% of customer service issues in the presence or absence of a proper FAQ and easy-to-use and powerful customer user interface?
techcrunch.com
This has been a big week for AI companies signing enterprise deals, with Zendesk unveiling new AI agents that are supposed to be able to resolve 80% of customer service issues, Anthropic and IBM announcing a strategic partnership, and Deloitte also announcing a deal with Anthropic.
Ready or not, enterprises are betting on AI | TechCrunch
This has been a big week for AI companies signing enterprise deals, with Zendesk unveiling new AI agents that are supposed to be able to resolve 80% of customer service issues, Anthropic and IBM announcing a strategic partnership, and Deloitte also announcing a deal with Anthropic.
techcrunch.com
michae.lv
Does your university have a contract with Grammarly? Write to the decision-maker asking if they think the university should be paying for a tool that is fast integrating features that can only be used for academic misconduct and cognitive offloading and request they drop the contract.
jedbrown.org
It is not "attribution and sourcing" to generate post-hoc citations that have not been read and did not inform the student's writing. Those should be regarded as fraudulent: artifacts testifying to human actions and thought that did not occur.
www.theverge.com/news/760508/...
For help with attribution and sourcing, Grammarly is releasing a citation finder agent that automatically generates correctly formatted citations backing up claims in a piece of writing, and an expert review agent that provides personalized, topic-specific feedback. Screenshot from Grammarly's demo of inserting a post-hoc citation.
https://www.grammarly.com/ai-agents/citation-finder

j2bryson.bsky.social
Why is @signal.org asking to see devices on my local network?

I'm used to (constantly) denying chrome that permission...

j2bryson.bsky.social
j2bryson.bsky.social
I wasted way to much time reading this this morning. There's nothing there except evidence that concentrated wealth/capital/leverage just disintegrates coherence. I got more out of a side trip into reading the Book of Daniel (NRSV) which is also incoherent but at least a bit poetic. 1/2
purgepalantir.bsky.social
"'I’m a libertarian, or a classical liberal, who deviates in one minor detail, where I’m worried about the antichrist,' Thiel said during his third lecture."

Okay, dude.

From The Guardian: www.theguardian.com/us-news/2025...