←
Arnd Roth200 posts
Arnd Roth avatar

Arnd Roth

@ArndRoth

Computational neuroscience and connectomics. Now at Bluesky

◷ Joined February 2021▧ 1 Photos and videos
Matt Angle avatar
⟳ Arnd Roth retweeted
Matt Angle@Matt_R_Angle·2h

Medtech is cool now. Listen to this podcast and tell me if you agree. @NEOM @PIF_en @paradromics

Telepathy Is Here: The BCI Revolution with Matt Angle, Paradromics and Babar Khan, NEOM Investment Fund. MEDTECH TELEPATHY MAY BE CLOSER THAN MOST PEOPLE REALIZE. To understand where BCI is really heading, I spoke with @Matt_R_Angle Founder and CEO of @paradromics, and Babar Khan, Head of Technology Investments at @NEOM Investment Fund. Matt described watching a participant think words internally and seeing those thoughts appear on a computer in real time. Babar called what he witnessed telepathy. But the conversation goes much deeper than one demonstration. We talked about why BCI could evolve from a medical device for restoring speech into a completely new computing platform connecting biology with machines. We covered the tradeoffs between invasive and non-invasive BCI, why signal quality and durability matter, and why AI cannot recreate neural information that a sensor never captured. We also discussed Paradromics' funding journey, including NIH and DARPA support, the regulatory and reimbursement bottlenecks holding back frontier MedTech, and why China's coordinated approach to BCI could make neurotechnology another major global technology race. Then we got into what may be the biggest long-term opportunity: brain data and AI. If AI agents could understand human intention directly through neural signals, the keyboard, mouse, touchscreen, and even voice may eventually become less important interfaces. We also explored consumer BCI, the Middle East's potential role in the industry, regional neural decoding, and Paradromics' ambition to build an ecosystem where other researchers and companies develop applications on top of its infrastructure. The first use case is restoring lost function. The bigger question is what happens when BCI becomes a platform for everything that comes next. Episode Breakdown: 00:00 How Brain-Computer Interfaces Could Change Humanity Forever 01:22 Meet Matt Angle and Babar Khan 02:10 What Is Paradromics and How Does Its BCI Technology Work? 03:40 NEOM's Investment Strategy for Frontier Technologies 05:43 Brain-Computer Interfaces Are Moving Beyond the Laboratory 06:46 Witnessing Real-Life Telepathy Through BCI 07:34 How Brain Implants Turn Thoughts Into Words 09:16 Why BCI Is an Engineering Challenge, Not Science Fiction 11:24 Is the 510(k) Pathway Holding Back MedTech Innovation? 12:54 Market Engineering and MedTech Growth 14:36 Why MedTech Has Been Thinking Too Small 16:10 Could BCI Companies Become More Valuable Than Intuitive Surgical? 17:59 Why Brain-Computer Interfaces Are More Than Medical Devices 19:15 What Really Drives Technology Platform Valuations? 19:52 Why the MedTech Industry Needs an Investment Bubble 21:10 How Breakthrough Medical Devices Could Overcome Reimbursement Barriers 22:06 Why Brain Implant Durability Matters 23:09 Implantable vs. Wearable Brain-Computer Interfaces 23:50 The Three Pillars of BCI: Safety, Durability and Performance 26:04 Is the Human Brain the Next Great Frontier? 26:58 Intracortical Brain Implants and Clinical Evidence 27:40 Why Recording Individual Neurons Matters 28:43 How BCI Could Unlock Human Potential 29:52 From Glucose Monitoring to Continuous Brain Monitoring 31:16 Could Brain Monitoring Improve Sleep, Stress and Mental Health? 32:33 How Complex Is Brain Implant Surgery? 33:14 Inside Paradromics' Brain Implant Technology 33:54 Making Brain Implant Surgery Accessible to More Hospitals 34:34 Paradromics vs. Neuralink: Is Robotic Surgery Necessary? 35:01 Why DeepTech Entrepreneurs Should Solve One Major Problem 36:11 China's Growing Lead in BCI Commercialization 37:02 The Global Race for Brain-Computer Interface Technology 37:53 What the US Can Learn From China's BCI Strategy 39:26 How Government Policy Could Accelerate MedTech Innovation 40:46 Why Regulatory Uncertainty Holds Back DeepTech Investment 42:46 How Paradromics Secured Its Early Funding 43:07 Inside Paradromics' $180 Million Funding Journey 43:48 The $14 Million DARPA Contract and Government Funding Trap 45:19 Why FDA and CMS Alignment Matters to MedTech Investors 46:23 Why Brain Data Could Power the Next Generation of AI 46:57 How Brain-Computer Interfaces Could Transform AI Agents 49:02 Elon Musk's Neuralink and the Value of Neural Data 50:05 Why Brain Data Could Become the Next Digital Gold 50:57 Could Your Thoughts Replace Keyboards and Touchscreens? 52:42 From Restoring Disabilities to Unlocking Human Superpowers 55:44 Could Brain-Computer Interfaces Become Consumer Technology? 57:45 The Scientific Limitations of Noninvasive Ultrasound BCI 58:08 Why AI Cannot Overcome the Laws of Physics 59:59 Why Capturing Brain Signals at the Source Matters 01:00:52 The 40-Bit Problem: Why Neural Data Quality Matters 01:02:02 When Does BCI Become the Next Computing Platform? 01:02:59 The Transition From Medical Devices to Consumer Brain Implants 01:04:01 Will Healthy People Get Brain Implants Within 10 Years? 01:04:47 Why NEOM Is Investing Beyond Traditional MedTech 01:07:25 How Paradromics Plans to Follow NVIDIA's Playbook 01:08:57 Open Hardware, Patents and the Future of BCI Innovation 01:09:54 Why Partnerships Will Shape the Future of MedTech 01:11:06 Outro

1:11:48
Anthropic avatar
⟳ Arnd Roth retweeted
Anthropic@AnthropicAI·Nov 6

New on our Science Blog: How AI is beginning to change scientific modeling. Scientific models help researchers explain and predict real-world systems—like how a single nerve cell fires, or what happens when neutron stars collide. Recent advances have transformed how scientists fit complex models to data. But scientists still typically write the models manually, and usually need to weigh up several competing versions, each of which might explain the data in a different way. In a guest post, Jakob Macke, Daniel Gedon, and Stefan Wahl describe how AI is starting to help with these parts of the process. By writing, running, fitting, and revising scientific models as code, AI can explore many possible model structures, and find out which explanations the data support. anthropic.com/research/teach…

Jacques Carolan avatar
⟳ Arnd Roth retweeted
Jacques Carolan@JacquesCarolan·Nov 5

Advances across neurotech (including today's chemogenetics!) have convinced me that powerful treatments for devastating brain conditions are possible. The next frontier is making them massively scalable, without requiring 10-hour brain surgery. Today, my team and I at @ARIA_research are announcing 10 teams we're backing with £50m to tackle this challenge. A thread 🧵 on their exciting projects!

Post attachment
Raphaël Millière avatar
⟳ Arnd Roth retweeted
Raphaël Millière@raphaelmilliere·Oct 16

No need to imagine! nature.com/articles/s41586-0… nature.com/articles/s41593-0… nature.com/articles/s41586-0… nature.com/articles/s41586-0… cell.com/neuron/abstract/S08… science.org/doi/10.1126/scia…

Imagine the roles reversed and you found AIs researching “mechanistic interpretability for humans”: Methods to interpret how individual human neurons relate to human behavior and reasoning, specific neurons you could stimulate to steer human behavior, methods to read human minds

Martin Bauer avatar
⟳ Arnd Roth retweeted
Martin Bauer@martinmbauer·Sep 26

This is the most relevant physics result achieved with AI yet It is derived in the most supersymmetric theory possible in 4D in the planar limit. This theory drowns so much in symmetry, it is finite and scale invariant. Basically the equivalent of a perfect sphere for QFTs

New on the Science Blog: Yes, Claude can do Nine Loops. Theoretical physicists predict how particles behave using formulas called scattering amplitudes. These are notoriously hard to compute, so researchers work with layers of increasingly fine corrections called “loops”—each added loop makes the answer more precise but takes exponentially more computation. Most calculations stop at two or three loops. Eight loops was the previous record in a simplified model physicists use as a testing ground (planar N=4 super-Yang-Mills), set by SLAC's Lance Dixon and collaborators. Last month, physicist and science writer @4gravitons issued a challenge: could an AI push past eight loops in this model, using only the compute budget an academic could reasonably access? Given a single prompt describing the nine-loop problem, Claude ran largely unsupervised for days in Claude Science and solved it using methods developed by Dixon and his colleagues, at a total cost of a few thousand dollars. Dixon independently verified the result, and von Hippel wrote about the experience for our blog. Read more: anthropic.com/research/yes-c…

Larry the Cat avatar
⟳ Arnd Roth retweeted
Larry the Cat@Number10cat·Sep 13

They claimed it was only illegal migrants who they had a problem with. They claimed it was only foreigners who couldn't speak English that they had a problem with They claimed it was only foreigners who didn't integrate that they had a problem with. Posts like this from a failed Reform candidate show that, actually, it's schoolchildren who have earned a place at high achieving schools that they have a problem with, if they happen to have black or brown skin. It doesn't matter to them if they came here legally, it doesn't matter to them if they have embraced life in this country, it doesn't matter if they were born here, to parents born here. They're not white so to warped individuals like Matthew, they're a problem. They aren't, people like him, and those that enable them are.

Post attachment
WEBKNOSSOS avatar
⟳ Arnd Roth retweeted
WEBKNOSSOS@webknossos·Sep 3

Got your data in WEBKNOSSOS and a coding agent like Claude? Turn it into a beautiful flythrough video. Let the AI write code to control camera moves, layer opacities and more, and show your audience what you actually see. Learn how in our blog post: home.webknossos.org/blog/cre…

0:15
Tobias Bonhoeffer avatar
⟳ Arnd Roth retweeted
Tobias Bonhoeffer@t_bonhoeffer·Aug 27

Check out our newest paper in bioRxiv, in which we provide strong causal evidence for the importance of dendritic spines in learning.  Phenomenal effort by @HiranmayJoag and great and very enjoyable collaboration with the Lüthi lab.

Excited to share our preprint, “Deleting learning-induced dendritic spines disrupts the memory they encode” biorxiv.org/content/10.64898… Joint effort with @t_bonhoeffer and Andreas Lüthi @Luthi_FMI and a big thanks to fantastic collaborators Patricio Opazo and @KentaM_Hagihara.

Arnd Roth avatar
Arnd Roth@ArndRoth·Aug 27

To be able to defend science against post-truth attacks from the right like climate change denial, we must ensure its integrity by calling out data fabrication and plagiarism, as in this case, and not punish the whistleblowers.

I think we have entered a weird post-truth moment in Britain, where it apparently doesn't matter if books and speaker bios are literally true: theatlantic.com/ideas/2026/0…

Arnd Roth avatar
Arnd Roth@ArndRoth·Aug 23

Freiheit ist immer Freiheit der Andersdenkenden (Rosa Luxemburg, 1918)

As a Chinese scientist, I find this episode deeply ironic. For decades, Western universities have presented themselves to the world—including China—as models of academic freedom. Yet an academic who publicly raised allegations of plagiarism has now been suspended while his university investigates his conduct. Whatever one thinks of @nathancofnas or his other views, the principle at stake should be simple: scholars must be free to scrutinize scholarship and present evidence of academic misconduct without fear of institutional retaliation. China has often been lectured by the West about freedom of thought. Perhaps it is time for Western academia to examine itself. Academic freedom means little if it protects only fashionable ideas and socially approved criticism. Its real test is whether it protects inconvenient inquiry, unpopular conclusions, and scholars whom the establishment dislikes. If you are an academic with an ORCID account, please consider signing the open letter: johnarmstrongmaths.com/openl…

Kanaka Rajan avatar
⟳ Arnd Roth retweeted
Kanaka Rajan@KanakaRajanPhD·Aug 7

Our approach for untangling brain-wide interactions is out today in @NeuroCellPress! Massive congrats to @mattperich for leading this incredible effort since 2020, and a huge thank you to all of our collaborators. doi.org/10.1016/j.neuron.202…

Post attachment

Excited to share our new @biorxivpreprint which introduces 𝘾𝙪𝙧𝙧𝙚𝙣𝙩-𝘽𝙖𝙨𝙚𝙙 𝘿𝙚𝙘𝙤𝙢𝙥𝙤𝙨𝙞𝙩𝙞𝙤𝙣 (𝘾𝙐𝙍𝘽𝘿)— an approach for inferring brain-wide interactions using data-constrained recurrent neural network (RNN) models. A thread 👇🏽 (1/8) biorxiv.org/content/10.1101/…

Demis Hassabis avatar
⟳ Arnd Roth retweeted
Demis Hassabis@demishassabis·Aug 5

I’ve been working towards AGI my whole life, and as we enter this pivotal moment, I’m stepping into a new role as Chair of Google DeepMind & Chief Scientist of Alphabet. This will allow me to focus on long-term strategy, and accelerating scientific breakthroughs, including leaning into my work at Isomorphic to help cure disease. I’m excited that @koraykv will be stepping up to lead GDM as SVP, alongside @joshwoodward and our exec team. I could not be more excited and confident about our amazing next chapter! 🚀 blog.google/company-news/ins…