Related posts

Co-founder of AngelList. Writes and podcasts at nav.al in short, aphoristic pieces on startups, wealth and judgement rather than long essays.

• Our readingkorrents.com

  Frontier AI labs improve via a flywheel of getting the smartest people on leading models, then distilling their inputs, model outputs, and synthetic data, with users locked in by mutual competition.

  Their words

  The Frontier Lab flywheel is to get the smartest people to use the leading models. Then distill their inputs, your own model’s outputs, and resulting synthetic data and environments. And the people have to use the leading models because they’re locked in mutual competition.

  @naval on Xx.com

Professor at Stanford working on robot learning and meta-learning; co-founded Physical Intelligence.

• Our readingkorrents.com

  Most future robotics training data will be robots' own autonomous attempts, the way most language-model data is now generated by the model itself.

  Their words

  Uh and just like how we see in language models how now a lot of time is spent actually generating data, generating synthetic data by actually running the model and having it think through things. I think a lot of the data in the future in robotics is going to be the robot attempting to do lots of tasks in lots of real world circumstances.

  youtube.com

Head of Instagram since 2018, previously head of Facebook's News Feed.

• Our readingkorrents.com

  An abundance of synthetic content will make people seek out the person behind the content more, not less.

  Their words

  but I think people are going to really seek out other points of view because Instagram was never just about the content. It was always about to a certain degree the person behind the content, the point of view, the reason they're sharing it, their perspective. And I think that's going to become more important, not less.

  youtube.com

Geneticist at Harvard; reads ancient DNA to work out how populations moved and mixed, and wrote Who We Are and How We Got Here.

• Our readingkorrents.com

  Ancient DNA has been a triumph about migration and a failure about biology, and the reason is simply that the samples were too small.

  Their words

  But, what's not been successful is learning about uh biology and biological change. And one big reason for that has been that the sample sizes have been too small.

  youtube.com

Co-founder of NVIDIA and, as of 2026, its chief executive; the company designs the GPUs most large AI models are trained on.

• Our readingkorrents.com

  Training is no longer limited by data but by compute, because most of the data models learn from is now synthetic.

  Their words

  The amount of data that we use to train models is going to continue to scale to the point where we're no longer limited… Training is no longer limited by… Data is now limited by compute. And the reason for that is most of the data is synthetic.

  youtube.com

Programmer, teacher and speaker; long-time Microsoft developer-community figure, host of the Hanselminutes podcast and author of the hanselman.com blog.

• Our readingkorrents.com

  AI-assisted coding will generate a lot of low-quality output that developers will need to clean up, even as it also produces good results.

  Their words

  I think that there will be lots of work for us cleaning up after the slop, but if you know what you're doing AI augmented development is going to get you some amazing results

  Is the craft dead?feeds.hanselman.com

Programmer, teacher and speaker; long-time Microsoft developer-community figure, host of the Hanselminutes podcast and author of the hanselman.com blog.

• Our readingkorrents.com

  Impressive one-shot AI coding demos are not a good way to judge AI coding ability.

  Their words

  People should not be judging AI-augmented coding by "1 shots."

  The danger of glamourizing one shotsfeeds.hanselman.com

Software developer and entrepreneur; co-founder of Heroku, author of The Twelve-Factor App, and a researcher at Ink & Switch.

• Our readingkorrents.com

  A personalized classifier needs several hundred labeled examples before its predictions become reliable.

  Their words

  Predictions were poor until I got to over 250 samples in the training set.

  Email triage with an embedding-based classifieradamwiggins.com

Founder of Terraform Industries and formerly a physicist at NASA's Jet Propulsion Laboratory; writes about energy, space and manufacturing.

• Our readingkorrents.com

  The United Kingdom is the one rich industrial country that cannot run on solar alone, and it has to build wind at scale instead.

  solarenergy

  Their words

  The only highly populated industrial country unable to trivially meet its electricity and synthetic fuel needs with solar alone is the United Kingdom, due primarily to a high population density and high latitude. The Nordic and Baltic countries are tiny by comparison. Among other problems, the UK needs to decide if it wants the future where energy is cheap and it is rich, or the future where energy is expensive and it is poor. If the former, it is time to get serious about large scale deployment of wind power, using home-grown vertically integrated technology at prices as low as $10/MWh.

  Energy Predictions 2025caseyhandmer.wordpress.com

Member of technical staff at Anthropic. He has led ML/AI teams at Amazon, Alibaba and Lazada, and writes about LLMs, recommender systems and engineering at eugeneyan.com.

• Our readingkorrents.com

  Human annotators can miss up to half of defects due to fatigue when reviewing large volumes of samples.

  Their words

  And human annotators can miss as many as 50% of the defects due to fatigue after looking at hundreds of samples.

  Product Evals in Three Simple Stepseugeneyan.com

Founding member of OpenAI and former director of AI at Tesla; creator of nanoGPT and the term "vibe coding".

From one piece Andrej Karpathy — “We’re summoning ghosts, not building animals” 2 beliefs, in the piece's order there

• Our readingkorrents.com

  Language models have no equivalent of sleep: nothing takes the day's context, thinks it over, and distils it back into the weights.

  Their words

  These models don't really have this distillation phase um of taking what happened, analyzing it, obsessively thinking through it, um basically doing some kind of a synthetic data generation process and distilling it back back into the weights

  youtube.com

• Our readingkorrents.com

  Every sample a model gives you is silently collapsed: each one looks fine, but together they occupy a tiny sliver of the possible space.

  Their words

  all of the samples you get from models are silently collapsed. They're silently, this is not obvious if you look at any individual example of it. They occupy a very tiny manifold of the possible space of um sort of thoughts about content.

  youtube.com

Reinforcement-learning researcher; co-wrote the field's standard textbook, wrote The Bitter Lesson, and shared the 2024 Turing Award.

• Our readingkorrents.com

  Succession to digital intelligence or to augmented humans is inevitable.

  Their words

  So I do think succession to digital or digital intelligence or augmented humans is inevitable.

  youtube.com

Founder of Terraform Industries and formerly a physicist at NASA's Jet Propulsion Laboratory; writes about energy, space and manufacturing.

• Our readingkorrents.com

  Nine tenths of the cost of Mars Sample Return comes from a single planetary-protection requirement.

  Their words

  something like 90% of the complexity and cost of MSR is driven by a requirement that the samples “break the chain” on the way home

  How to get samples back from Marscaseyhandmer.wordpress.com

Professor at Berkeley working on robot learning; co-founded Physical Intelligence, and argues that robots will learn from data rather than be programmed.

• Our readingkorrents.com

  Synthetic experience can only rehearse what a system already knows; information about the world has to be injected from outside.

  Their words

  So here's what I would say that deep down at a very fundamental level the synthetic experience that you create yourself doesn't allow you to learn more about the world. It allows you to rehearse things. It allows you to consider counterfactuals but somehow information about the world needs to get injected into the system.

  youtube.com

Founder of Terraform Industries and formerly a physicist at NASA's Jet Propulsion Laboratory; writes about energy, space and manufacturing.

• Our readingkorrents.com

  Cheap synthetic fuel made from sunlight and air would help China more than America, because it frees China from the oil it has to import.

  ChinaAmerica

  Their words

  It absolutely asymmetrically helps China. We're not currently working with China. They don't plan to, but like the physics is very obvious, you know, and synthetic fuels have been around for 100 years.

  youtube.com

Co-founder and CEO of Google DeepMind, and a 2024 Nobel laureate in Chemistry for the protein-structure work behind AlphaFold.

• Our readingkorrents.com

  AI will not run out of data, because there is already enough real-world data to build the simulators that generate more of it.

  Their words

  Do you have enough data to make simulations, so that you can create more synthetic data that are from the right distribution? Obviously that's the key. So you need enough real-world data in order to be able to create those kinds of data generators, and I think that we're at that step at the moment.

  youtube.com

Structural engineer and senior infrastructure fellow at the Institute for Progress; writes Construction Physics on why physical things cost what they cost.

• Our readingkorrents.com

  Synthetic quartz could replace mined high-purity quartz at five to ten times the price, which chipmakers could absorb.

  Their words

  The downside is that synthetic quartz is much more expensive, perhaps on the order of 5–10 times the cost of natural high-purity quartz.

  construction-physics.com

Machine learning engineer and independent AI consultant. He writes about LLM evaluation, tooling and applied ML at hamel.dev, and previously worked on machine learning at GitHub.

• Lovedrcmnd.app

  Hex

  Their words

  You might be skeptical of using synthetic data. After all, it’s not real data, so how can it be a good proxy? In my experience, it works surprisingly well. Some of my favorite AI products, like Hex use synthetic data to power their evals

  hamel.dev

Biologist at the MIT Media Lab; invented CRISPR gene drives and then spent years arguing for safeguards against their misuse and against engineered pandemics.

From one piece It shouldn't be easy to buy synthetic DNA fragments to recreate the 1918 flu virus 3 beliefs, in the piece's order there

• Our readingkorrents.com

  Ordering the synthetic DNA needed to rebuild the 1918 influenza virus is surprisingly easy, even from companies that screen customers' orders.

  Their words

  It should be hard — exceedingly hard — to obtain the synthetic DNA needed to recreate the virus that caused the deadly 1918 influenza pandemic without authorization. But my lab found that it’s surprisingly easy, even when ordering gene fragments from companies that check customers’ orders to detect hazardous sequences.

  media.mit.edu

• Our readingkorrents.com

  The genome sequences of pandemic viruses and the protocols for building them from synthetic DNA are already freely available online.

  Their words

  Both the genome sequences of pandemic viruses and step-by-step protocols to make infectious samples from synthetic DNA are now freely available online.

  media.mit.edu

• Our readingkorrents.com

  Every synthetic DNA order should be screened for hazardous sequences and shipped only to researchers whose work a biosafety authority has approved.

  Their words

  That makes it essential to ensure that all synthetic DNA orders are screened to determine whether they contain hazardous sequences, which should be shipped only to legitimate researchers whose work has been approved by a biosafety authority.

  media.mit.edu

Research scientist at Meta in Berlin working on multilingual models and evaluation; led the multilingual team at Cohere and was a research scientist at Google DeepMind before that.

• Our readingkorrents.com

  Retrieval-augmented generation is the most practical approach for reducing LLM hallucinations.

  LLMs

  Their words

  Retrieval-augmented generation (RAG; Lewis et al., 2020), which conditions on the LLM's generation on retrieved documents is the most practical paradigm IMO.

  Command R+ruder.io

Writer and illustrator behind Wait But Why, a long-form blog of stick-figure-illustrated explainers on science, technology and society, and author of the book What's Our Problem?

• Our readingkorrents.com

  Augmented reality and virtual reality are no longer separate categories, but points on one continuum.

  Their words

  And it's a general reminder that AR and VR2 being separate categories is a thing of the past.

  All My Thoughts After 40 Hours in the Vision Prowaitbutwhy.com

Founder of Platformer, a newsletter on tech platforms and the people they affect, and co-host of the Hard Fork podcast at the New York Times.

• Recommendsrcmnd.app

  DALL-E 3

  Their words

  I suggest using a text-to-image generator like DALL-E 3. However you feel about synthetic media, the pace of improvement is readily apparent from a single prompt

  x.com

Designer, architect and scientist, who led the Mediated Matter group at MIT and founded OXMAN, a company that builds products in collaboration with living organisms.

• Our readingkorrents.com

  The runaway emergence we are trying to prevent in AGI is exactly what we should be trying to cause in synthetic biology and bio-based materials.

  AGI

  Their words

  What we don't want to happen with AGI, we want to happen with synthetic biology. What we don't want to happen online and software with language, we want for it to happen with bio-based materials.

  youtube.com

Journalist; wrote We Are Eating the Earth on food and climate, and writes the newsletter Emissions.

• Our readingkorrents.com

  Synthetic fertiliser is a serious climate and pollution problem and also the reason half the people alive are alive.

  Their words

  They suck! But they also feed the world. Half the people alive today would be very not alive if fertilizers manufactured from natural gas weren’t super-sizing our harvests.

  Farming Needs to Suck Lessmichaelgrunwald.substack.com