j2bryson.bsky.social
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
4,441 posts
8,572 followers
523 following
Regular Contributor
Active Commenter
comment in response to
post
Oh
comment in response to
post
Wow, linkedin is not talking much about this (or seeing it) even though I didn't mention Microsoft explicitly, maybe they parse the linked articles too? www.linkedin.com/feed/update/... #shaddowbans are so insidious because you can never tell if it's your post or the platform. #dsa
comment in response to
post
Here you go...
www.computerweekly.com/opinion/Micr...
comment in response to
post
#agi and #aiethics contribute to this — these narratives and technologies are shaping the socioeconomic and sociopolitical, psychological landscape which they also weaponise into a battlefield.
Zelenskyy is an absolute hero.
2/2
#globsec2025 #udhr #humanrights #humancentring @hertiecfr.bsky.social
comment in response to
post
I wish everyone at the meeting could read this blogpost – kind of whiplash to go from worrying about sentience to worrying about how easy it is to use LLM like bad regular expressions! simonwillison.net/2025/Jun/11/...
comment in response to
post
Despite @simonwillison.net usual propensity for anthropomorphism, that entire post makes it blindingly obvious how not far from regular expressions LLMs really are (but less reliable.)
Quite a change from the @ruhr-uni-bochum.de meeting the previous two days...
comment in response to
post
That series was my favourite thing on TV all week. Better than STtNG. I forgot how much I looked forward to it.
comment in response to
post
Open the pdf with word, then export as docx
comment in response to
post
They expect AGI in 3.
Also, Europe has been producing more AI PhDs for some time. And coincidentally mostly don’t anticipate AGI, except the UK, so are still building useable products.
comment in response to
post
June has 31 days?
comment in response to
post
I find footnotes just an extra jump I have to make. Some popular books now have parallel citations as endnotes with absolutely no notation in the main text; we're considering this for my book. What I want is e-books letting you flip between reference style. Toggle to see how/if claims are supported.
comment in response to
post
@lovingautomaton.bsky.social weird dunno why it wasn't clickable before.
comment in response to
post
@lovingautomaton.bsky.social is here too!
comment in response to
post
Or using the present moment to up their surveillance game.
#AIEthics
comment in response to
post
and she said "If AI does have goals & sentience we under recognise then they might have reasons to behave in ways unaligned with our interests." I hadn't realised that Big Tech was actively using #anthropomorphism to THREATEN governments / regulators.
#AGI #ArtificialConsciousness #AIConsciousness
comment in response to
post
AFAIK tend to focus all kinds of tricky situationsm, moral or not. One of my friends kept rehearsing a terrible skiing accident, but anyway science presently conceives of dreams as helping consolidate memory (integrate experience into concepts / skills.) Do you attend to dilemmas a lot when awake?
comment in response to
post
and she said "If AI does have goals & sentience we under recognise then they might have reasons to behave in ways unaligned with our interests." I hadn't realised that Big Tech was actively using #anthropomorphism to THREATEN governments / regulators.
#AGI #ArtificialConsciousness #AIConsciousness
comment in response to
post
Was just wondering whether to ask a Googler whether people "just happened" to be talking about moral status of AI increasingly, or vested interests might be driving that, #AGI #AIConsciousness 1/2
comment in response to
post
none yet. Way more excited about bsky.app/profile/j2br...
comment in response to
post
@markgbaxter.bsky.social see above re that grant we could never get funded, cf www.nature.com/articles/nat... and www.cell.com/neuron/fullt... from the Halassa lab.
comment in response to
post
People are talking a lot about this bsky.app/profile/anil...
comment in response to
post
We determine whose well being when we design things. This isn't about waiting to see, it's about adequate engineering and regulation, like any other output of commerce.
comment in response to
post
The first speaker works for a company focussed on "AI sentience and well being"
This may be a rough couple of days. More on @j2bryson.mastodon.social.ap.brid.gy