🌱NEW PREPRINT & SPECIAL day! We unveil an advance for plant science, eng, and synbio in first (of many) Khalil <> Gehring collabs, led by the one-and-only @willmshaw. Thread below on principles design of a new, single-integrating T-DNA vector for precision plant xformation! 👇

The Spurs know how to TAKE the lead, they just don't know how to HOLD the lead. And that's really the most important part of the lead: the holding

Found research from May that shows e. coli "remember" environmental conditions their ancestors lived through and behave differently because of it, all without a brain, neurons, or a nervous system. Inherited proteins handle the memory. The mechanism is what's so cool. Different cellular components relax on different timescales, so the cell integrates its history as a power law instead of a single decay constant. This is an extremely elegant expression of memory as a biochemical hidden state, emergent from ribosome dynamics, that lets a granddaughter cell respond to stress it never personally experienced. It is also oddly close to the logic behind recurrent neural networks. My main takeaways are that memory might predate minds and a growing belief that AI is convergent with biological intelligence. paper: journals.aps.org/prxlife/abs…

Who did it better

Cast of Laocoön and his Sons (Roman version of a lost Greek original), 100BC-50AD

Did @qedScience step in at half time. Give the half time team talk?? Impressive !

Is OG god?

It will take years to fully grasp the philosophical and moral implications of this win, and OG's heroics. It proves that man is fundamentally good, and that anything is possible.

The Knicks are going to win tonight. @qedScience, is that a valid claim?

"The Case for a Validity Layer in Life Science Research" - a very interesting article by Meir Maor, QED's CTO. Meir explains why biology is harder for AI than mathematics.

AI is taking away some of our cognitive skills (and that's natural, just like we lost the ability to spell without mistakes). It will be especially sad to lose some of these, like understanding other languages. AI will probably make speaking other languages a luxury, but speaking another language is not just about communication, it's also about different ways of thinking. The same is definitely true for writing (a skill that may be halfway toward extinction). A lot of the time, I don't know what I think until I write it down. Of course, since I am developing an AI for scientists (QED science), I think about this all the time, and that's why I am not so excited about AI that "thinks" for scientists. We're trying to build the AI that helps scientists think.

Pretty interesting story in @ScienceMagazine this week on what looks like a serious problem in the senescence field. More than 400 papers apparently used the wrong antibody for p16-INK4a — an antibody that actually recognizes a completely different, unrelated protein (a component of the actin cytoskeleton). This affects work on senescent cell accumulation in aging and disease, and most critically, some of the evidence base for senolytic drug research. What concerns me most is that many of these papers somehow got the "right" answer using the wrong antibody. That's not just an innocent reagent mix-up — it raises real questions about data fabrication or selective reporting in at least some of these labs. I've commented before about how ignoring data that doesn't fit the narrative is a major problem in certain areas of the longevity literature (e.g. sirtuins and NAD), and here a potentially widespread example in senescence. Hopefully journals will investigate and retract as necessary, but based on my experience that seems optimistic. One concrete fix is that journals should flag problematic antibody product codes at submission so reviewers can catch this before publication. Reviewers should absolutely be on the lookout for this going forward. However, these fixes won't address the larger problem. We need to understand how these scientists got the results they wanted and published them over 400 times (!!!): whether through intentional deception, incompetence, accident, or some legitimate explanation. Credit for discovering this goes to @addictedtoigno1 who wrote about it first on his blog: For Better Science science.org/content/article/…?

AI labs wanting to solve biology and learning how our data is organized

Instead of a robot scientist, what we really want is a robot that sits in the audience and nods in agreement

Validation in biology isn't just hard for AI. It's hard for everyone. Peer reviewer agreement on accept/reject is poor across domains. The Amgen study found only 6 of 53 landmark cancer papers were reproducible. AI agents reasoning on that literature inherit every flaw directly. The infrastructure gap isn't only about data access. It's about having no reliable way to evaluate what the science actually means. That's the problem we're working on at @qedScience

Relevant during the NBA playoffs: White nights kill you (but catching up the next night fixes it)

Has there ever been a controlled study comparing the good cop/bad cop method with good cop/bad cop/indifferent cop?

Some people on psychedelics report seeing consciousness everywhere. For example Michael Pollan in his new book about consciousness "A world appears" describes how under the influence of LSD (if i remember correctly) plants appeared conscious to him. I wonder whether on LSD people chatting with AI would suddenly be convinced that it's conscious (someone should do an experiment!)

It's a very tense situation and no one knows what will happen next. However, No team came back from 0-2 at home in the Finals. Go Knicks!

How your email finds me