Elizabeth Gibney
@lizziegibney.bsky.social
Senior reporter at Nature, views my own. Journalist covering physics, AI, policy. Attempting to stop lurking and start posting.
See my stories at nature.com/news
See my stories at nature.com/news
There was a lot of hubbub last month about this tiny, one-author model that beats some big LLMs on a prominent benchmark supposed to test intelligence (ARC-AGI).
Right now it mainly does sudokus, but could this technique eventually helps AI go beyond LLMs?
www.nature.com/articles/d41...
🧪🤖
Right now it mainly does sudokus, but could this technique eventually helps AI go beyond LLMs?
www.nature.com/articles/d41...
🧪🤖
'Tiny' AI model beats massive LLMs at logic test
Technique could be used as a cheap way to boost ability of other AI models.
www.nature.com
November 13, 2025 at 4:03 PM
There was a lot of hubbub last month about this tiny, one-author model that beats some big LLMs on a prominent benchmark supposed to test intelligence (ARC-AGI).
Right now it mainly does sudokus, but could this technique eventually helps AI go beyond LLMs?
www.nature.com/articles/d41...
🧪🤖
Right now it mainly does sudokus, but could this technique eventually helps AI go beyond LLMs?
www.nature.com/articles/d41...
🧪🤖
This is pretty cool. How do you train an LLM? Yes but reeeeally how? This 200+ page (!) blog from @hf.co shows how to train a model from start to finish, bugs, warts & all. Loads of interesting details I hadn't thought about huggingface.co/spaces/Huggi...
huggingface.co
October 31, 2025 at 11:30 AM
This is pretty cool. How do you train an LLM? Yes but reeeeally how? This 200+ page (!) blog from @hf.co shows how to train a model from start to finish, bugs, warts & all. Loads of interesting details I hadn't thought about huggingface.co/spaces/Huggi...
Another day, another claim of quantum advantage -- this time, with hints of (someday) doing something useful 🧪⚛️
www.nature.com/articles/d41...
www.nature.com/articles/d41...
Google claims ‘quantum advantage’ again — but researchers are sceptical
The firm says it has solved a problem on a quantum processor faster than a classical computer, and is optimistic about future scientific applications.
www.nature.com
October 23, 2025 at 9:06 AM
Another day, another claim of quantum advantage -- this time, with hints of (someday) doing something useful 🧪⚛️
www.nature.com/articles/d41...
www.nature.com/articles/d41...
Where do the world famous LLM sceptic @garymarcus.bsky.social & his bud Laurence Fishburne (AKA Morpheus) like to hang out? The Royal Society of course! 🧪🤖
Here's my write up from @unisouthampton.bsky.social's Celebrating the 75th Anniversary of the Turing Test event www.nature.com/articles/d41...
Here's my write up from @unisouthampton.bsky.social's Celebrating the 75th Anniversary of the Turing Test event www.nature.com/articles/d41...
AI language models killed the Turing test: do we even need a replacement?
Chatbots now ace the mathematician's famous imitation game, but imitation never equalled intelligence.
www.nature.com
October 21, 2025 at 6:20 AM
Where do the world famous LLM sceptic @garymarcus.bsky.social & his bud Laurence Fishburne (AKA Morpheus) like to hang out? The Royal Society of course! 🧪🤖
Here's my write up from @unisouthampton.bsky.social's Celebrating the 75th Anniversary of the Turing Test event www.nature.com/articles/d41...
Here's my write up from @unisouthampton.bsky.social's Celebrating the 75th Anniversary of the Turing Test event www.nature.com/articles/d41...
Next week @jameszou.bsky.social & colleagues will host a conference where all the papers are written by AI agents & reviewed by them too.
What do you reckon? A good chance to put AIs through their paces? Or a way to divert AI slop from elsewhere? 🧪🤖
My story here:
www.nature.com/articles/d41...
What do you reckon? A good chance to put AIs through their paces? Or a way to divert AI slop from elsewhere? 🧪🤖
My story here:
www.nature.com/articles/d41...
AI bots wrote and reviewed all papers at this conference
Event will assess how reviews by models compare with those written by humans.
www.nature.com
October 15, 2025 at 11:00 AM
Next week @jameszou.bsky.social & colleagues will host a conference where all the papers are written by AI agents & reviewed by them too.
What do you reckon? A good chance to put AIs through their paces? Or a way to divert AI slop from elsewhere? 🧪🤖
My story here:
www.nature.com/articles/d41...
What do you reckon? A good chance to put AIs through their paces? Or a way to divert AI slop from elsewhere? 🧪🤖
My story here:
www.nature.com/articles/d41...
Reposted by Elizabeth Gibney
Not common to see a 7-byline story in the wild, but that's what the situation calls for when the government shuts down and every science agency needs to be checked in on for RIFs, grant terminations, and general dysfunction.
Our update on what the chaos means for science here:
Our update on what the chaos means for science here:
Scientists lose jobs and grants as US government shutdown takes a toll
Hundreds of people at the US Centers for Disease Control and Prevention have received layoff notices, and work at many federal laboratories has been suspended.
www.nature.com
October 15, 2025 at 1:59 AM
Not common to see a 7-byline story in the wild, but that's what the situation calls for when the government shuts down and every science agency needs to be checked in on for RIFs, grant terminations, and general dysfunction.
Our update on what the chaos means for science here:
Our update on what the chaos means for science here:
I'm pretty sceptical about 'AI scheming', as it's so easy to anthropomorphise & experiments often involve telling the AI to do the bad thing they end up doing.
To understand what's behind the hype, read this smart & sober overview from@silverjacket.bsky.social 🧪🤖
www.nature.com/articles/d41...
To understand what's behind the hype, read this smart & sober overview from@silverjacket.bsky.social 🧪🤖
www.nature.com/articles/d41...
AI models that lie, cheat and plot murder: how dangerous are LLMs really?
Tests of large language models reveal that they can behave in deceptive and potentially harmful ways. What does this mean for the future?
www.nature.com
October 9, 2025 at 10:59 AM
I'm pretty sceptical about 'AI scheming', as it's so easy to anthropomorphise & experiments often involve telling the AI to do the bad thing they end up doing.
To understand what's behind the hype, read this smart & sober overview from@silverjacket.bsky.social 🧪🤖
www.nature.com/articles/d41...
To understand what's behind the hype, read this smart & sober overview from@silverjacket.bsky.social 🧪🤖
www.nature.com/articles/d41...
Over at @naturepodcast.bsky.social we also made a super short vid on this year's physics Nobel Prize. Quantum phenomena at the macroscopic scale in two minutes... go! With me, and Ben Thompson & camera work by the fab @emzywb.bsky.social 🧪⚛️
www.youtube.com/shorts/krita...
www.youtube.com/shorts/krita...
October 8, 2025 at 11:16 AM
Over at @naturepodcast.bsky.social we also made a super short vid on this year's physics Nobel Prize. Quantum phenomena at the macroscopic scale in two minutes... go! With me, and Ben Thompson & camera work by the fab @emzywb.bsky.social 🧪⚛️
www.youtube.com/shorts/krita...
www.youtube.com/shorts/krita...
Reposted by Elizabeth Gibney
John Martinis’ wife didn’t wake him in the middle of the night (California time) to tell him he had won a Nobel. “I got up a little bit before 6. Then I opened my computer and saw John and Michel’s and my pictures."
Story by @lizziegibney.bsky.social and me
www.nature.com/articles/d41...
Story by @lizziegibney.bsky.social and me
www.nature.com/articles/d41...
Groundbreaking quantum-tunnelling experiments win physics Nobel
John Clarke, Michel Devoret and John Martinis discovered quantum physics on a macroscopic scale, paving the way for quantum computing.
www.nature.com
October 7, 2025 at 4:30 PM
John Martinis’ wife didn’t wake him in the middle of the night (California time) to tell him he had won a Nobel. “I got up a little bit before 6. Then I opened my computer and saw John and Michel’s and my pictures."
Story by @lizziegibney.bsky.social and me
www.nature.com/articles/d41...
Story by @lizziegibney.bsky.social and me
www.nature.com/articles/d41...
Have you heard loads about "AI agents", but have little idea what they are, what they can do for researchers and whether to believe the hype? Then this is for you! 🧪🤖https://www.nature.com/articles/d41586-025-03246-7
A scientist’s guide to AI agents — how could they help your research?
Researchers are increasingly turning to artificial-intelligence tools that can handle complex, multi-step processes.
www.nature.com
October 6, 2025 at 8:22 AM
Have you heard loads about "AI agents", but have little idea what they are, what they can do for researchers and whether to believe the hype? Then this is for you! 🧪🤖https://www.nature.com/articles/d41586-025-03246-7
Nature's news team is now recruiting for our paid US intern position for January to June 2026. It's a great opportunity to come join our team & publish important stories 🧪
Apply by 16 Oct springernature.wd3.myworkdayjobs.com/SpringerNatu...
Apply by 16 Oct springernature.wd3.myworkdayjobs.com/SpringerNatu...
News Intern, Nature News
Job Title: News Intern, Nature News Location: Washington, D.C or New York — Hybrid Working Application Deadline: October 15, 2025 About Springer Nature Springer Nature is one of the leading publishers...
springernature.wd3.myworkdayjobs.com
September 29, 2025 at 8:17 AM
Nature's news team is now recruiting for our paid US intern position for January to June 2026. It's a great opportunity to come join our team & publish important stories 🧪
Apply by 16 Oct springernature.wd3.myworkdayjobs.com/SpringerNatu...
Apply by 16 Oct springernature.wd3.myworkdayjobs.com/SpringerNatu...
I was going to post a rant about the perils of quantum computing hype, following the claim by #HSBC & #IBM that they have improved bond market predictions using a QC, but instead I will just link to Scott Aaronson's blog which says it all scottaaronson.blog?p=9170 🧪⚛️ cc @bullshitquantum.bsky.social
HSBC unleashes yet another “qombie”: a zombie claim of quantum advantage that isn’t
Today, I got email after email asking me to comment on a new paper from HSBC—yes, the bank—together with IBM. The paper claims to use a quantum computer to get a 34% advantage in predic…
scottaaronson.blog
September 26, 2025 at 11:35 AM
I was going to post a rant about the perils of quantum computing hype, following the claim by #HSBC & #IBM that they have improved bond market predictions using a QC, but instead I will just link to Scott Aaronson's blog which says it all scottaaronson.blog?p=9170 🧪⚛️ cc @bullshitquantum.bsky.social
Also read @helenpearson.bsky.social & @heidiledford.bsky.social's excellent story unpicking the origins of Trump's paracetamol/autism claims www.nature.com/articles/d41...
This quote sums it up: "We do not think that taking acetaminophen is in any way contributing to actually causing autism"
This quote sums it up: "We do not think that taking acetaminophen is in any way contributing to actually causing autism"
September 23, 2025 at 11:27 AM
Also read @helenpearson.bsky.social & @heidiledford.bsky.social's excellent story unpicking the origins of Trump's paracetamol/autism claims www.nature.com/articles/d41...
This quote sums it up: "We do not think that taking acetaminophen is in any way contributing to actually causing autism"
This quote sums it up: "We do not think that taking acetaminophen is in any way contributing to actually causing autism"
Remember DeepSeek's R1 model that crashed the US stock market in Jan? DeepSeek has said it did not boost the model by training on OpenAI outputs. This and much more (eg $$ to train & technical details) revealed in the firm's peer reviewed paper out in Nature today 🧪🤖 www.nature.com/articles/d41...
Secrets of DeepSeek AI model revealed in landmark paper
First peer reviewed study shows how a Chinese start-up firm made the market-shaking LLM for US$300,000.
www.nature.com
September 17, 2025 at 7:10 PM
Remember DeepSeek's R1 model that crashed the US stock market in Jan? DeepSeek has said it did not boost the model by training on OpenAI outputs. This and much more (eg $$ to train & technical details) revealed in the firm's peer reviewed paper out in Nature today 🧪🤖 www.nature.com/articles/d41...
ICYMI, this week I wrote about GPT-5 and the effort to stop LLMs from hallucinating, as well as saying they've done tasks they haven't. With web search to verify facts, finding citations is getting a lot better. But as soon as it's off, things can still go very wonky🧪🤖 www.nature.com/articles/d41...
Can researchers stop AI making up citations?
OpenAI’s GPT-5 hallucinates less than previous models do, but cutting hallucination completely might prove impossible.
www.nature.com
September 11, 2025 at 9:11 AM
ICYMI, this week I wrote about GPT-5 and the effort to stop LLMs from hallucinating, as well as saying they've done tasks they haven't. With web search to verify facts, finding citations is getting a lot better. But as soon as it's off, things can still go very wonky🧪🤖 www.nature.com/articles/d41...
OpenAI's new open-weight #AI model looks to be a powerful reasoner. It's small enough to use locally and they've done loads to make it available.
But has it entered the game too late to become the go-to for researchers? Do ping me if you're trying it out!
www.nature.com/articles/d41... 🤖🧪
But has it entered the game too late to become the go-to for researchers? Do ping me if you're trying it out!
www.nature.com/articles/d41... 🤖🧪
OpenAI launches reasoning LLM that you can download and tweak
One version of the gpt-oss large language model can run on a laptop, and performs nearly as well as the company’s most powerful models.
www.nature.com
August 7, 2025 at 10:55 AM
OpenAI's new open-weight #AI model looks to be a powerful reasoner. It's small enough to use locally and they've done loads to make it available.
But has it entered the game too late to become the go-to for researchers? Do ping me if you're trying it out!
www.nature.com/articles/d41... 🤖🧪
But has it entered the game too late to become the go-to for researchers? Do ping me if you're trying it out!
www.nature.com/articles/d41... 🤖🧪
Hey quantum fans. Here's a slightly different #quantum survey from my mind-boggling one on interpretations, but seems important! This group hopes to understand how to help quantum scientists to stay in the field & thrive. If you have time, you can fill it out here👇 www.diviq.org/survey 🧪⚛️
Global Quantum Needs Assessment Survey — Diversity In Quantum
www.diviq.org
August 5, 2025 at 3:29 PM
Hey quantum fans. Here's a slightly different #quantum survey from my mind-boggling one on interpretations, but seems important! This group hopes to understand how to help quantum scientists to stay in the field & thrive. If you have time, you can fill it out here👇 www.diviq.org/survey 🧪⚛️
This year quantum physics turns 100, so @nature.com decided to conduct the biggest ever survey of what lies behind it. Do physicists really believe in multiple universes? Can influences happen instantaneously? Is the Copenhagen interpretation all it's cut out to be? ⚛️🧪 www.nature.com/articles/d41...
Physicists disagree wildly on what quantum mechanics says about reality, Nature survey shows
First major attempt to chart researchers’ views finds interpretations in conflict.
www.nature.com
July 30, 2025 at 2:19 PM
This year quantum physics turns 100, so @nature.com decided to conduct the biggest ever survey of what lies behind it. Do physicists really believe in multiple universes? Can influences happen instantaneously? Is the Copenhagen interpretation all it's cut out to be? ⚛️🧪 www.nature.com/articles/d41...
The Thirty Meter Telescope — long planned to be built in Hawaii but now with its US funding on the chopping block — could be given a new lease of life in Spain, following a government bid to host at the telescope on La Palma. Will the TMT board accept?
🧪🔭 www.nature.com/articles/d41...
🧪🔭 www.nature.com/articles/d41...
Spain bids €400 million to host mega telescope at risk in US budget cuts
New Canary Islands home could save controversial Thirty Meter Telescope first proposed for Hawaii.
www.nature.com
July 26, 2025 at 5:43 PM
The Thirty Meter Telescope — long planned to be built in Hawaii but now with its US funding on the chopping block — could be given a new lease of life in Spain, following a government bid to host at the telescope on La Palma. Will the TMT board accept?
🧪🔭 www.nature.com/articles/d41...
🧪🔭 www.nature.com/articles/d41...
Moonshot AI has now published its technical report about Kimi K2
github.com/MoonshotAI/K...
The 32-page doc says the model trained on NVIDIA H800 GPUs, including post-training reinforcement learning on agentic examples
My story from last week is here: www.nature.com/articles/d41...
github.com/MoonshotAI/K...
The 32-page doc says the model trained on NVIDIA H800 GPUs, including post-training reinforcement learning on agentic examples
My story from last week is here: www.nature.com/articles/d41...
github.com
July 22, 2025 at 12:11 PM
Moonshot AI has now published its technical report about Kimi K2
github.com/MoonshotAI/K...
The 32-page doc says the model trained on NVIDIA H800 GPUs, including post-training reinforcement learning on agentic examples
My story from last week is here: www.nature.com/articles/d41...
github.com/MoonshotAI/K...
The 32-page doc says the model trained on NVIDIA H800 GPUs, including post-training reinforcement learning on agentic examples
My story from last week is here: www.nature.com/articles/d41...
Meet Kimi K2 from Beijing-based Moonlight AI. Good at coding, writing & multi-tool tasks, reports say it's the best open model available.
It looks like #DeepSeek wasn't an anomaly. Sources said to expect more cutting-edge & free models from Chinese firms soon 🤖🧪
www.nature.com/articles/d41...
It looks like #DeepSeek wasn't an anomaly. Sources said to expect more cutting-edge & free models from Chinese firms soon 🤖🧪
www.nature.com/articles/d41...
‘Another DeepSeek moment’: Chinese AI model Kimi K2 stirs excitement
The latest version of the chatbot, developed by start-up Moonshot AI, is open for researchers to build on.
www.nature.com
July 17, 2025 at 9:21 AM
Meet Kimi K2 from Beijing-based Moonlight AI. Good at coding, writing & multi-tool tasks, reports say it's the best open model available.
It looks like #DeepSeek wasn't an anomaly. Sources said to expect more cutting-edge & free models from Chinese firms soon 🤖🧪
www.nature.com/articles/d41...
It looks like #DeepSeek wasn't an anomaly. Sources said to expect more cutting-edge & free models from Chinese firms soon 🤖🧪
www.nature.com/articles/d41...
For this week's @naturepodcast.bsky.social we chatted about a news story on researchers putting hidden prompts into preprints to try to game peer review, and turned it into a 90-sec vid.
Take a look here: www.youtube.com/shorts/WvamK... 🧪🤖
With @nickpetrichowe.bsky.social & @danjfox.bsky.social
Take a look here: www.youtube.com/shorts/WvamK... 🧪🤖
With @nickpetrichowe.bsky.social & @danjfox.bsky.social
Could hidden AI prompts game peer review?
YouTube video by Nature Podcast
www.youtube.com
July 16, 2025 at 4:18 PM
For this week's @naturepodcast.bsky.social we chatted about a news story on researchers putting hidden prompts into preprints to try to game peer review, and turned it into a 90-sec vid.
Take a look here: www.youtube.com/shorts/WvamK... 🧪🤖
With @nickpetrichowe.bsky.social & @danjfox.bsky.social
Take a look here: www.youtube.com/shorts/WvamK... 🧪🤖
With @nickpetrichowe.bsky.social & @danjfox.bsky.social
"POSITIVE REVIEWS ONLY"
Researchers are hiding instructions for AI peer reviewers in their preprints. I'm almost impressed by how much some of these researchers hide within a tiny white space.
But it's far from clear how much of an impact such prompts have.
www.nature.com/articles/d41...
⚛️🧪🤖
Researchers are hiding instructions for AI peer reviewers in their preprints. I'm almost impressed by how much some of these researchers hide within a tiny white space.
But it's far from clear how much of an impact such prompts have.
www.nature.com/articles/d41...
⚛️🧪🤖
July 14, 2025 at 9:41 AM
"POSITIVE REVIEWS ONLY"
Researchers are hiding instructions for AI peer reviewers in their preprints. I'm almost impressed by how much some of these researchers hide within a tiny white space.
But it's far from clear how much of an impact such prompts have.
www.nature.com/articles/d41...
⚛️🧪🤖
Researchers are hiding instructions for AI peer reviewers in their preprints. I'm almost impressed by how much some of these researchers hide within a tiny white space.
But it's far from clear how much of an impact such prompts have.
www.nature.com/articles/d41...
⚛️🧪🤖
Thanks so much to the @absw.bsky.social for my award, and for the fantastic event last night. It was wonderful to hang out with such an excellent bunch of journalists!
Congrats to the other winners & nominees.
I will cherish my #bestmugintheworld forever!
Congrats to the other winners & nominees.
I will cherish my #bestmugintheworld forever!
🏆#ABSWawards 2025
👉Research Policy or Funding Story of the Year
🎉Elizabeth Gibney @lizziegibney.bsky.social @springernature.com
👏👏👏Congratulations!!!
👉Research Policy or Funding Story of the Year
🎉Elizabeth Gibney @lizziegibney.bsky.social @springernature.com
👏👏👏Congratulations!!!
July 10, 2025 at 9:01 AM
Thanks so much to the @absw.bsky.social for my award, and for the fantastic event last night. It was wonderful to hang out with such an excellent bunch of journalists!
Congrats to the other winners & nominees.
I will cherish my #bestmugintheworld forever!
Congrats to the other winners & nominees.
I will cherish my #bestmugintheworld forever!