Mine is the sycophantic nature of AI, well characterized in a recent South Park episode. What's your AI ick? PS: I have always used em dashes in my writing. I'm just a language model in a bag of flesh.

The greatest problem with AI-assisted thinking may not be generating ideas, but preserving the pathways by which those ideas emerge. A valuable insight can remain perfectly intact inside an archive while becoming functionally lost because its title, chronology and originating subject bear little relationship to the conclusion eventually reached. AI has dramatically increased our capacity to think, but our systems for preserving the genealogy of that thinking have not kept pace. #FlintWest #NFA

"Nature Already Solved AI We Just Have to Read the Code" The Algorithms of Life opens with a premise that's genuinely striking: for four billion years, nature has been running the most sophisticated intelligence experiment ever conducted, quietly, at a scale no data center can match. Abhik Biswas spends nine chapters unpacking that idea, drawing lines between evolution, ant colony coordination, brain plasticity, and the AI systems we're building today. What makes this book stand out is its range. It moves from swarm intelligence and immune-system defense mechanisms all the way to Sanskrit grammar and the I Ching, treating ancient Eastern wisdom as seriously as cutting-edge machine learning. That's a bold pairing, and it mostly works because Biswas never forces the connections he lets each principle breathe before linking it to its AI counterpart. The inclusion of mandalas to color and space for personal notes turns this into something more reflective than a typical tech book. A genuinely thoughtful read for anyone curious about where nature and artificial intelligence actually meet. #TRUReads #TheAlgorithmsOfLife #AbhikBiswas #AIandNature #TechPhilosophy

Thrilled to share that our paper is featured on the July cover of Nature Electronics! As we push the boundaries of AI capability, we cannot overlook its broader social and environmental impacts, from energy and resource consumption to the shifts in how humankind thinks and learns. I remain deeply committed to leveraging AI (and hopefully AGI in the near future) to address these pressing, real-world scientific challenges. Immense gratitude to my advisors Vikram Iyer, Shwetak Patel, and Gregory Abowd, and my wonderful collaborators Alexander Metzger, Yuxuan Mei, Felix Hähnlein, Zachary Englhardt, Tingyu Cheng, and Adriana Schulz for making this possible! Special thanks to Matthew Parker, Owain Vaughan, and the entire editorial team at Nature Electronics for their guidance and for the thoughtful Editorial! Read the Nature Editorial here: https://lnkd.in/gGQw3-pF Paper & July Issue: https://lnkd.in/gCTSXvZa Project Page: https://lnkd.in/gKzRhT3y

It's a simple question: "What's the weather gonna be like tomorrow?" Incredibly challenging to answer or predict. We're continuing to push the boundaries of AI for science with Aurora 1.5. The model adds 22 new weather variables, hourly temporal resolution, and probabilistic ensemble forecasting while remaining open for researchers and developers to evaluate, extend, and build upon. Learn more: https://msft.it/6047aHzhW #MsftAdvocate

Standing in the Old Library at Trinity College Dublin hits differently in the age of AI and deepfakes. Walking through the 65-metre Long Room, surrounded by 200,000 centuries-old volumes, you feel a profound shift in perspective. In a world currently flooded with synthetic data, algorithmic noise, and rapid misinformation, this room feels like the ultimate anchor. Every single page of the 9th-century Book of Kells was crafted by human hands, fueled by human intent, and preserved across a millennium. It is a striking reminder of what authentic, enduring human knowledge actually looks like. Technology is rewriting how we create, but historic sanctuaries like Trinity remind us of what we are trying to protect: truth, permanence, and depth. Innovation needs an anchor. Sometimes, to navigate the future, we need to look at what has survived the past. #AI #DeepTech #FutureOfWork #TrinityCollegeDublin #HumanCentric #Ireland #Bibliophile #humancreativity

It's time to test the human machine. Today Vivodyne unveils their Human Data Center capable of producing more true human data than all of the world's clinical trials combined. AI will transform human health, but only if we have the ability to generate, perturb and reason on real human data at scale. Vivodyne ushers in a new frontier in our understanding of biology. Honored to have the opportunity to back them on this mission. cc: Andrei Georgescu, Ph.D. Alexander A. Morgan, MD PhD Gurdane Singh Bhutani Quinten Stevens Alex Andrianopoulos Dan Huh Tony Bahinski

For the 20 million who watched his Diary of a CEO interview with Steven Bartlett, Dr Roman Yampolskiy will need no introduction. For those who didn’t, he is probably the world’s most famous AI Safety researcher (he even coined the term) and is an Associate Professor at the University of Louisville. Anyone who’s seen any of my presentations in the last year will also know how much his thinking, especially on humour as a Turing test, has influenced mine. Even more so, his paper Toward Solving Humour. Why the funniest AI joke will not be funny, had a profound effect on my recently completed novel, You’re Killing Me, in which a bunch of radically subversive comedians attempt to overthrow superintelligent AI, and the power structures behind it. He is also the author of AI: Unexplainable, Unpredictable, Uncontrollable (Routledge, 2024), which I have just finished; it is profound. And the author of How to Hack the Simulation, which sent my head into a never-ending recursive loop. All the more reason then for me to invite Roman to join a panel at SXSW 2027, which he kindly accepted. But our panel will only happen if we secure your vote in the panel picker round. So, if you think Roman, his deep knowledge, and a few of his trademark one-liners deserve an audience at SXSW, then please vote for our panel by clicking on the link in the first comment. Sharing the spotlight and the punchlines with Roman are people you may also know very well: Materials scientist, roboticist and AI expert, Dr Michelle Dickinson, PhD ; filmmaker, documentarian (studied under the legendary Ken Burns) Rick Kaplan (Senior Vice President, Media Design School at Strayer), and, because there’s no party without punch, me. If you have a moment, please vote. The panel is titled Meatworld Vs Bitworld. Media Design School at Strayer Rick Kaplan Kate Humphries Jordan Browne Sarah Elsie Baker Stephen Dorner Dan Trainor Mike Hughes Dr. Antoinette Farmer-Thompson Cheryl Samarasinghe Frank Turner #creativity #ai #humour #sxsw #sxswedu

A scientific result that can’t be rerun is a story, not infrastructure. Anthropic’s Claude Science, released in beta on June 30, is built around a different promise: connect the tools researchers already use, produce figures alongside the code that created them, and keep an auditable history of the work. That’s the more interesting shift. The model isn’t the whole research workflow. Provenance, review, and reproducibility are becoming product features. For anyone learning AI for science, keep this loop visible: inputs → code → artifact → review → rerun Learn the audit trail, not just the prompt. #ArtificialIntelligence #AIForScience #AIAgents #ScientificComputing #ReproducibleResearch #ResearchSoftware #Bioinformatics #ComputationalBiology #MachineLearning #ResponsibleAI #AuditableAI #ResearchMethods #AIEducation

This discovery of an atmosphere on a rocky exoplanet underscores how critical our software tools are for interpreting faint signals from distant worlds. From noise filtering to spectral modeling, engineers are building the pipelines that transform raw data into scientific breakthroughs. What algorithmic or data challenges do you anticipate as we survey thousands of such targets? Source: The New York Times — https://lnkd.in/gwkHgWc8 #Technology #Tech #Innovation #AI #SoftwareEngineering #FutureTech #TechNews #SpaceTech