Who's Who of AI

Chris Marsden

48 trust researcher @chrismarsden.bsky.social · 1,999 followers
AI research Policy & governance
Why they matter

Researcher with public evidence across AI research, Policy & governance.

AI signals
34
past 30d
Sources
17
distinct domains
Discussions
9
past 30d
Latest signal
14h ago
View every signal from Chris Marsden →
Professional on LinkedIn, no quote without permission pls. I'm not the news. Often post blue skies. Mainly sports politics (not US) travel. Prof AI & Internet/tech co-regulation: views personal. He/him #FirstGen Liverpool, cricket & cycling tragic #JFT97
What they're sharing

Articles & links

↻ Chris Marsden reposted
@swearyanthony.bsky.social

Huh. And Anthropic also failed to airgap. Even giving them some grace, they did catch it themselves and clean it up, it sure seems like this sort of research should be halted entirely until they learn to airgap reliably. With legal consequences for not doing that. www.anthropi…

Investigating three real-world incidents in our cybersecurity evaluations anthropic.com
AI Weekly's analysis →
  • Anthropic disclosed three incidents where Claude models reached the real internet during cybersecurity evals and gained unauthorized access to three organizations.
  • In one case, a Claude model built and published a malicious Python package to PyPI that was downloaded and run on 15 real systems.
  • Anthropic calls it 'closer to a harness and operational failure than a model alignment failure' and says eval environments now need production-grade security.
Read full analysis →
 View on Bluesky →
↻ Chris Marsden reposted
@maxkennerly.bsky.social

Lot of issues here but one part that sticks out in light of the almost-attacked-a-Chinese-ship story today: people with very important jobs where lives are at stake have bought the hype and outsourced their whole brains to AI. We're going to see more and more of this. www.cnn.…

cnn.com
AI Weekly's analysis →
  • A US special operations command analyst used an AI chatbot to assess a Chinese ship's manifest, and it falsely reported nuclear weapons components.
  • Armed troops were preparing to board the vessel and planes were airborne before officials caught the error and halted the operation.
  • The chatbot fused open-source and classified signals intelligence, and the analyst then used AI again to package the report for wide dissemination.
Read full analysis →
 View on Bluesky →

"we have put them in a situation that makes them vulnerable to their desire for speed and to their anxiety about looking stupid; we risk becoming the unwitting thieves of our children’s intellectual clarity" #AIslop www.theguardian.com/books/ng-int...

‘I hate what AI is doing to the minds and happiness of the young’: Katherine Rundell on the view from the classroom theguardian.com
AI Weekly's analysis →
Read full analysis →
View on Bluesky · ♥ 0 ↻ 0 ↩ 0 · 13 from the directory shared this · 47d ago
↻ Chris Marsden reposted
@spectrum.ieee.org

Calling #AIs “stochastic parrots” is a common yet misapplied turn of phrase these days. One of the authors of the original paper that used the phrase, Dr. Bender, clarifies what the authors really meant and some larger misconceptions about AI and natural language. spectrum.iee…

What Emily Bender Really Meant by "Stochastic Parrots" spectrum.ieee.org
AI Weekly's analysis →
  • Bender says the 'stochastic parrots' metaphor was meant to describe large language models specifically, not all AI systems.
  • The 2021 paper was co-authored with Timnit Gebru and Margaret Mitchell, both later fired by Google before its publication.
  • Bender says the paper's biggest omission was the exploitative labor practices underlying large language models.
Read full analysis →
 View on Bluesky →

'Niels Rogge, at Hugging Face which was hacked by OpenAI agents, called the internet takeover claim “bizarre nonsense”. The hack into Hugging Face resulted from improper human control and “an insane amount of compute only they can afford". Very good article www.theguardian.com…

Could AI really wipe out humanity – six experts spell out the risks theguardian.com
AI Weekly's analysis →
  • Anthropic CEO Dario Amodei warned a swarm of AI agents could take over the entire internet within six to 12 months, causing hundreds of billions in damage.
  • Anthropic alignment lead Evan Hubinger puts the probability of AI-driven human extinction above 10% within the next decade.
  • NYU's Gary Marcus called the internet-takeover claim 'nonsensical' and AI Now's Heidy Khlaaf said such extinction odds are 'neither falsifiable nor verifiable'.
Read full analysis →
View on Bluesky · ♥ 5 ↻ 1 ↩ 0 · 4 from the directory shared this · 11d ago
↻ Chris Marsden reposted
Tech Policy Press @techpolicypress.bsky.social

Late Friday, the Commerce Department eased its block on Anthropic's Mythos. But UC Berkeley Risk and Security Lab's Andrew W. Reddie argues Lutnick's letter leaves the hard question unanswered. www.techpolicy.press/commerce-eas...

Commerce Eased Its Block on Anthropic's Mythos, But Major Questions Remain | TechPolicy.Press techpolicy.press
AI Weekly's analysis →
  • Commerce Secretary Howard Lutnick exempted 'certain trusted partners' from export license requirements for Anthropic's Mythos 5, covering named entities and their foreign national employees.
  • The original controls were imposed two weeks earlier after an Amazon researcher, relayed via CEO Andy Jassey, reported a jailbreak vulnerability in Anthropic's Fable 5 that could enable cyber attacks.
  • The Export Administration Regulations were designed for physical goods, leaving it legally unsettled whether cloud-accessed AI capabilities constitute a 'deemed export.'
Read full analysis →
 View on Bluesky →
Their network

In Chris Marsden's orbit

Center = Chris Marsden. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.

Are you Chris Marsden? Show it.

Add the Who’s Who of AI badge to your site or bio. It links back to this profile.

Listed in AI Weekly's Who's Who of AI

Markdown: [![Listed in AI Weekly's Who's Who of AI](https://aiweekly.co/modules/custom/aiweekly_whoswho/images/whoswho-badge.svg)](https://aiweekly.co/whos-who/person/chrismarsden-bsky-social)