Elliot Jones
@elliotmjones.bsky.social
Research Scientist @ UK AI Security Institute
Reposted by Elliot Jones
Today (w/ @ox.ac.uk @stanford @MIT @LSE) we’re sharing the results of the largest AI persuasion experiments to date: 76k participants, 19 LLMs, 707 political issues.
We examine “levers” of AI persuasion: model scale, post-training, prompting, personalization, & more!
🧵:
We examine “levers” of AI persuasion: model scale, post-training, prompting, personalization, & more!
🧵:
July 21, 2025 at 4:20 PM
Today (w/ @ox.ac.uk @stanford @MIT @LSE) we’re sharing the results of the largest AI persuasion experiments to date: 76k participants, 19 LLMs, 707 political issues.
We examine “levers” of AI persuasion: model scale, post-training, prompting, personalization, & more!
🧵:
We examine “levers” of AI persuasion: model scale, post-training, prompting, personalization, & more!
🧵:
Reposted by Elliot Jones
We're hiring a Research Engineer for the Societal Resilience team at the AI Security Institute. The role involves building data pipelines, web scraping, ML engineering, and creating simulations to monitor these developments as they happen.
job-boards.eu.greenhouse.io/aisi/jobs/46...
job-boards.eu.greenhouse.io/aisi/jobs/46...
Research Engineer - Societal Resilience
London, UK
job-boards.eu.greenhouse.io
June 9, 2025 at 6:17 PM
We're hiring a Research Engineer for the Societal Resilience team at the AI Security Institute. The role involves building data pipelines, web scraping, ML engineering, and creating simulations to monitor these developments as they happen.
job-boards.eu.greenhouse.io/aisi/jobs/46...
job-boards.eu.greenhouse.io/aisi/jobs/46...
Reposted by Elliot Jones
This does mean one extremely exciting thing - 🚨Job Alert Klaxon🚨
Ada is hiring a new Associate Director to lead our Emerging Tech and Practice research team. I urge you to apply - you'll get to work with some of the best researchers on high impact projects.
app.beapplied.com/apply/kefx3r...
Ada is hiring a new Associate Director to lead our Emerging Tech and Practice research team. I urge you to apply - you'll get to work with some of the best researchers on high impact projects.
app.beapplied.com/apply/kefx3r...
Associate Director, Emerging Technology and Industry Practice - Ada Lovelace Institute
The Ada Lovelace Institute (Ada) is a hiring an Associate Director to lead our Emerging Technology & Industry Practice research directorate and collectively set its agenda and workplan in our next...
app.beapplied.com
March 20, 2025 at 8:49 AM
This does mean one extremely exciting thing - 🚨Job Alert Klaxon🚨
Ada is hiring a new Associate Director to lead our Emerging Tech and Practice research team. I urge you to apply - you'll get to work with some of the best researchers on high impact projects.
app.beapplied.com/apply/kefx3r...
Ada is hiring a new Associate Director to lead our Emerging Tech and Practice research team. I urge you to apply - you'll get to work with some of the best researchers on high impact projects.
app.beapplied.com/apply/kefx3r...
Reposted by Elliot Jones
Capable of executing tasks and providing conversation or advice, Advanced AI assistants are highly personable and personalised.
Our policy briefing defines what makes an Advanced AI Assistant, how they work and how they differ from other forms of AI.
www.adalovelaceinstitute.org/policy-brief...
Our policy briefing defines what makes an Advanced AI Assistant, how they work and how they differ from other forms of AI.
www.adalovelaceinstitute.org/policy-brief...
Delegation Nation
Advanced AI Assistants and why they matter
www.adalovelaceinstitute.org
February 4, 2025 at 2:38 PM
Capable of executing tasks and providing conversation or advice, Advanced AI assistants are highly personable and personalised.
Our policy briefing defines what makes an Advanced AI Assistant, how they work and how they differ from other forms of AI.
www.adalovelaceinstitute.org/policy-brief...
Our policy briefing defines what makes an Advanced AI Assistant, how they work and how they differ from other forms of AI.
www.adalovelaceinstitute.org/policy-brief...
Reposted by Elliot Jones
AI companions are on the rise, but what are the possible long-term effects on people and society?
@jamiebernardi.bsky.social examines the potential benefits and risks - both individual and systemic - of this type of AI service.
www.adalovelaceinstitute.org/blog/ai-comp...
@jamiebernardi.bsky.social examines the potential benefits and risks - both individual and systemic - of this type of AI service.
www.adalovelaceinstitute.org/blog/ai-comp...
Friends for sale: the rise and risks of AI companions
What are the possible long-term effects of AI companions on individuals and society?
www.adalovelaceinstitute.org
January 24, 2025 at 2:02 PM
AI companions are on the rise, but what are the possible long-term effects on people and society?
@jamiebernardi.bsky.social examines the potential benefits and risks - both individual and systemic - of this type of AI service.
www.adalovelaceinstitute.org/blog/ai-comp...
@jamiebernardi.bsky.social examines the potential benefits and risks - both individual and systemic - of this type of AI service.
www.adalovelaceinstitute.org/blog/ai-comp...
Reposted by Elliot Jones
Yesterday was all about AI Opportunities...
We're expecting a frontier AI Bill...
But what plans does govt need for the harms & challenges already with us?
@connectedbydata.org, Demos & @adalovelaceinst.bsky.social convened civil society/industry for some answers
demos.co.uk/research/the...
We're expecting a frontier AI Bill...
But what plans does govt need for the harms & challenges already with us?
@connectedbydata.org, Demos & @adalovelaceinst.bsky.social convened civil society/industry for some answers
demos.co.uk/research/the...
‘The Frontier AI Bill and…?’: Briefing Paper
Demos is Britain’s leading cross-party think-tank. We produce original research, publish innovative thinkers and host thought-provoking events.
demos.co.uk
January 14, 2025 at 11:07 AM
Yesterday was all about AI Opportunities...
We're expecting a frontier AI Bill...
But what plans does govt need for the harms & challenges already with us?
@connectedbydata.org, Demos & @adalovelaceinst.bsky.social convened civil society/industry for some answers
demos.co.uk/research/the...
We're expecting a frontier AI Bill...
But what plans does govt need for the harms & challenges already with us?
@connectedbydata.org, Demos & @adalovelaceinst.bsky.social convened civil society/industry for some answers
demos.co.uk/research/the...
Reposted by Elliot Jones
A window into the blackbox
- longer form thoughts on new latest batch of algorithms published by the UK Government
medium.com/@imogencathe... @adalovelaceinst.bsky.social and ht @halcyene.bsky.social
- longer form thoughts on new latest batch of algorithms published by the UK Government
medium.com/@imogencathe... @adalovelaceinst.bsky.social and ht @halcyene.bsky.social
A window into the black box: new entries on the UK government’s AI register
The Government has just published a new batch of records describing how algorithms are being used in the public sector.
medium.com
December 18, 2024 at 11:33 AM
A window into the blackbox
- longer form thoughts on new latest batch of algorithms published by the UK Government
medium.com/@imogencathe... @adalovelaceinst.bsky.social and ht @halcyene.bsky.social
- longer form thoughts on new latest batch of algorithms published by the UK Government
medium.com/@imogencathe... @adalovelaceinst.bsky.social and ht @halcyene.bsky.social
Reposted by Elliot Jones
EU Codes of Practice for GPAI models are out - and it's a great draft. Very impressed by the working groups and team who've driven this.
A few thoughts.
digital-strategy.ec.europa.eu/en/library/f...
A few thoughts.
digital-strategy.ec.europa.eu/en/library/f...
First Draft of the General-Purpose AI Code of Practice published, written by independent experts
Independent experts present the first draft of the General-Purpose AI Code of Practice, which will be discussed with around 1000 stakeholders next week. The AI Office is facilitating relevant AI Act u...
digital-strategy.ec.europa.eu
November 14, 2024 at 2:10 PM
EU Codes of Practice for GPAI models are out - and it's a great draft. Very impressed by the working groups and team who've driven this.
A few thoughts.
digital-strategy.ec.europa.eu/en/library/f...
A few thoughts.
digital-strategy.ec.europa.eu/en/library/f...
Reposted by Elliot Jones
an excerpt from "The Unaccountability Machine", in which I try to outline a theory of the sort of thing that keeps happening
open.substack.com/pub/backofmi...
open.substack.com/pub/backofmi...
the only message the channel can carry is a scream
my diagnosis
open.substack.com
November 6, 2024 at 12:26 PM
an excerpt from "The Unaccountability Machine", in which I try to outline a theory of the sort of thing that keeps happening
open.substack.com/pub/backofmi...
open.substack.com/pub/backofmi...