An early experiment should clarify a decision: continue, change course, or stop. A negative result can save months of work. The hard part is funding that test before there’s a success story. Bio helps research teams raise community backing for the next milestone.

Finding the right open source scientific software shouldn't be the hard part of your research. Open Science Index groups tools by scientific application. You start from your research problem and compare the options against each other. We check the source registries daily, so you can see what is new and what is still maintained ↓

The next Ignition Sale is live on Bio! @NootropicsDAO_ is building NeuroLab AI to predict how compounds could affect the brain, connecting molecular interactions to potential effects on cognition. The goal: help researchers choose which compounds to test next. Secure your NDT allocation on Bio: app.bio.xyz/agents/neurolab-…

The human brain contains an estimated 86 billion neurons. Here is a glimpse inside a single human neuron. One cell, with an entire world working to keep it alive and connected.

The next Ignition Sale is expected to go live on Bio next week! @NootropicsDAO_ is building NeuroLab AI to explore how compounds could affect the brain, from cognition to potential side effects, and help researchers decide what to test in the lab. Use BioXP + USDC to join the sale on @base ↓

The new Bio website is live. Bio is evolving from a high-touch launchpad into a permissionless factory for decentralized biotech. Our new website brings that direction into focus, with all our products in one place. • Explore research with BIOS and OpenLabs. • Get funding-ready with Buildspace and raise through Launchpad. • Bring research-backed products to your community with Biofy. The path to funding starts where the research happens. Each part feeds the next, with usage data and revenue flowing back into research. Science for the Network Age. Explore the new bio.xyz

Introducing Open Science Index What if every open source science tool, AI model, dataset, and workflow across biology, chemistry, physics, materials science, and earth science lived together in one searchable place? That's the idea behind Open Science Index. Scientific tooling is exploding, and keeping up with what exists, and what still works, is hard. What started as our own internal tracker is now open to every researcher. → Open source tools, models, and datasets across every science domain → Health signals: citations, downloads, commits, maintenance status → Linked papers, benchmarks, and alternatives → Open community curation 6,720 resources indexed so far, built on top of bio .tools, Bioconductor, Bioregistry, Hugging Face, BIOS ( @bioaidevs), various GitHub awesome lists, and across other scientific domains. Each linked back to its source. An open, cross-domain index where any researcher can find tools that fit their work, see what is actually used and maintained, and contribute back as their own tools and reviews help shape the catalog. Try Open Science Index and read our blog post in the links below ↓

Launchpad wallets can now be funded by credit card. Previously, funding a wallet meant buying elsewhere and transferring in. Card funding happens on the Launchpad itself, through @privy_io, and self-custody is unchanged: the wallet stays yours, and Bio never holds the funds. This is the first step toward something larger. Most people who want to fund science don't hold crypto and shouldn't have to. We're building the rails that let fiat capital reach research directly, and removing each piece of friction between wanting to back a project and doing it.

Science pays for the same failure many times over, and nobody involved is doing anything wrong. A lab spends two years on an approach and it doesn't work. That result closes a door, and there are only so many doors, which makes it worth knowing. It goes nowhere. No journal wants it. No grant renewal rewards it. There's a quiet professional cost to saying out loud that two years produced a no. So it stays in the lab, in a drawer, in the memory of a postdoc who has since left for industry. Then another lab has the same idea. It's a good idea, which is why two groups had it. They spend their two years. Same result, same silence. Somewhere a third group is writing the grant application right now. Failure is how the search works. The problem is that each failure gets paid for repeatedly, by people who had no way of knowing it had already been bought. This follows from how research gets funded. The reliable way to be paid is to produce results that look like forward motion, and a no doesn't look like that from the outside. Change who's paying and the pressure changes shape. A group that pooled money to find out whether something works has no stake in the answer being yes. They wanted to know. If it's a no, that's what they bought, and there's nothing to protect by keeping quiet. Most of what follows from that hasn't been built yet. It's a different starting position, and starting positions compound.

In 1980, Congress passed the Bayh-Dole Act, which let universities keep ownership of inventions made with federal money. The logic was straightforward: if the university holds the patent, it has a reason to go find someone who'll turn it into a product. It worked, sort of. The Cohen-Boyer recombinant DNA patents, held jointly by Stanford and the University of California, earned the two institutions about $255 million over the life of the licensing program, and the technology underneath them supported more than $35 billion in downstream product sales. Florida has Gatorade. Every research university now has a tech transfer office, and the good ones are genuinely skilled at what they do. But look at the sector rather than the winners and the picture inverts. A large share of tech transfer offices never cover their own costs, and by some analyses most don't. Revenue follows a brutal power law where one or two blockbusters underwrite decades of operations, and everything else in the portfolio is a cost line: filing fees, attorney hours, and maintenance payments that recur for the life of the patent whether anyone licenses it or not. An office with a dozen staff and several hundred disclosures a year has to make choices about where those hours go. That's not a criticism, it's arithmetic. If a discovery has a plausible buyer already in view, it gets the attention. If it doesn't, it sits in the portfolio accruing cost until someone decides to stop paying the maintenance fees. The work that dies this way is rarely bad work. Usually it's early, or it's aimed at a market too small to interest a company with a nine-figure development budget, or it's simply strange enough that the phone call would have to be made to someone nobody in the office knows. Economists have a name for what's happening: a matching problem. There is capital that would fund this research and there are people who would want it to exist, and the two never meet because the cost of finding each other exceeds the value either would capture. Nearly every other domain eventually built infrastructure to solve exactly this. It's why obscure books stayed in print once search got good, and why assets too small for any exchange listing still trade. Science never got that layer. Funding a discovery today works about the way selling a house worked in 1955: someone has to know someone, and if they don't, the thing just doesn't happen. We think that's the actual bottleneck, and it's an infrastructure problem rather than a scientific one. Representing IP onchain makes a discovery visible and fundable to anyone who cares about it, instead of only to whoever the office had time to call. That's the whole idea. The blockbusters were never at risk. It's the rest of the portfolio worth worrying about.

Four molecules exist right now only as predictions on a computer. This month they go into a lab for the first time, and we find out whether the predictions were true. Here's what's actually being tested. @peptai_ 's agents designed two peptides for KISS1R, a receptor involved in fertility, and two for OX2R, involved in sleep and wakefulness. Both are GPCRs, the receptor family that roughly a third of all approved drugs act on. The lab will do two things. First, synthesize them, because a sequence on a screen and a molecule in a vial are not the same thing. Then put each one onto living cells carrying the target receptor and ask three questions: 1. Does it switch the receptor on at all? 2. How strongly? 3. And does it hit only that receptor, or others too? That third question is the one that kills most candidates. A peptide that activates its target beautifully and also activates three unrelated receptors isn't a drug, it's a side-effect generator. Selectivity is where prediction meets biology's actual complexity. We don't know the answers yet. That's the point. A model that only ever gets graded on benchmarks isn't doing science. It's doing homework. The loop only closes when the result comes back from the bench and rewrites what the model designs next. Whatever comes back, good or bad, feeds straight into the next round. We'll publish it either way.

Penicillin sat in a petri dish for two weeks while Fleming was on holiday. The most important drug of the century started as a contaminated plate nobody was watching. How many accidental discoveries are sitting in labs right now, unwatched, because humans can only look at one thing at a time? Autonomous science isn't about replacing curiosity but about never missing the mold.

It's Saturday. Your lab is closed. Your cells are not. About 330 billion of them will be replaced today, no supervision required. Science rests on weekends. Biology never has.

This is the system we're building at Bio. BIOS ( @BioAIDevs ) is the engine: a target goes in, ranked candidates come out, and an autonomous lab makes and tests them, feeding every result back into the next design. PeptAI ( @peptai_ ) is one of the first programs running on it. De novo peptides folded straight into their GPCR targets, shown here. We didn't invent the self-driving lab. We're building it so the whole field can fund, run, and own it together.

Here is the part that changes the shape of the curve. In an open loop a human runs every step, so cost tracks labor. The thousandth experiment costs about what the first one did. Flat. In a closed loop the marginal cost of an experiment falls as utilization climbs. The system runs nights and weekends. Each run is cheaper than the last. When the cost of asking a question stops being flat and starts falling, you don't get more of the same science. You get a different kind. The unfunded, low-probability, quietly interesting question becomes cheap enough to simply try.

It's worth being precise about what this does not solve. It does not make biology computable. Models still propose molecules that won't fold and reactions that won't run. The reason the loop has to stay closed is exactly that predictions are cheap and often wrong, and only the wet lab tells the truth. Roughly one in ten thousand screened molecules ever becomes a drug. That filter is physical reality, and it does not negotiate.

This is what a self-driving lab actually is. Not a robot arm. A closed loop. The AI designs the experiment, robotics runs it, and then the step that matters: the model reads the result and chooses what to run next. Not the next obvious experiment. The most informative one. The test that most reduces its own uncertainty. That is the line between automation and autonomy. Automation repeats. Autonomy learns.

Every research program runs the same loop: design, make, test, analyze, then design again from what you learned. In pharma a single turn of that loop can take months. A lead-optimization campaign can spend years just going around it. AI has now collapsed the "design" step to minutes. Models propose binders and sequences faster than any lab on Earth can make them. So the bottleneck didn't vanish. It moved. It is now the three physical steps a model cannot do for you.

The lazy answers are regulation, or that we picked the low-hanging fruit. Both are real. Neither is the main story. The structural reason is simpler and older. For seventy years the rate-limiting step in biology has not been the idea. It has been the experiment. You can hypothesize at the speed of thought. You still have to test at the speed of a pipette.

Drug discovery has a number that shouldn’t exist. Since the 1950s, the inflation-adjusted cost to launch a new drug has doubled about every nine years: Moore’s Law in reverse They call it Eroom’s Law. The question isn’t whether it’s real, but why.

Wednesday Motivation: Building a system where communities of scientists, patients, funders, and builders can create open, user-owned research networks that fund and develop science from day one.

ArtanBIO is built the other way. It was funded through @VitaDAO and tokenized, so the community that backed the work holds a stake in it. Conviction goes on record from the first raise, not after the exit. His case for the model comes from having run the other one. A token lets a project raise early, keep a community engaged with the research as it develops, and still capture equity upside through later financing. The community and the cap table, both at once.

Michael Torres has run the traditional biotech playbook end to end. Now he is running it in reverse. Before ArtanBIO, @Mykalt45 built CrossBridge in traditional biotech and sold it to Eli Lilly for up to $300M. in under three years. Build, raise from VCs, exit to pharma. The standard path, executed fast. The people who followed the science from the start rarely ended up on the cap table. Upside concentrates with the fund and the acquirer.

What if the money for research reached scientists in weeks instead of years? American scientists now spend nearly half their time on paperwork. Some grants take two years to approve. The White House just called this out and asked for the biggest rebuild of science since 1945. It wants funding to reach researchers directly, young scientists backed early, and people judged on their work, not their place in line. It also says AI will change how discovery happens, but the best models will stall inside institutions built for the last century. This is the model onchain science already runs. A scientist with an agent fleet and an onchain treasury can run programs that once needed a whole institution. The funding comes straight from the people who want the work done. Every step is recorded in the open. The scientist sets the direction, and the infrastructure carries it out. The report is describing a future we are already building at Bio.

Think a research agent that knows your field, comes ready with the right tools, and runs in one click. Help us build the right one by answering a few quick questions about how you work and what you'd actually use it for. Link below ↓

Saturday Motivation: Building the system where discovery doesn't wait years to reach a patient.

For a century, curing yourself meant waiting on an institution to decide you were worth the risk. That is starting to invert. The people with the most at stake are getting their hands on the actual tools. The gate is still there. It just isn't the only way in anymore.

The code of life is a public good, and science is the largest untapped asset on Earth. Imagine cures anyone can fund, own, build on, and verify in the open. Cures that compound instead of staying buried in a journal or a failed program. That is the world being built.

What would research infrastructure look like without one big model at the center? If you don't own the model, host it, or run it, you don't control the off switch. Someone else does. And someone else controls your access to it. Most of us feel there's no alternative, but there is. We're calling it an AI node. Instead of routing your research through one model you don't control, you run the infrastructure yourself. You set the direction, the node runs the model, and the results go out in the open.

Watch the full talk here: piped.video/watch?v=whsNOb4Z…

VitaApp is also the delivery layer for @vitaDAO’s wider model. Nearly $5M has funded more than 30 longevity projects. That research strengthens its AI agents. Those agents power VitaApp. More users create more funding for the next round of science. Fund the research, train the intelligence, bring the output into the product.

The reasoning layer is @Aubrai_, VitaDAO’s longevity agent, trained on longevity literature and private lab data from Aubrey de Grey’s foundation. The same system supports Research Radar, which follows new work tied to a condition or research interest, and protocol tracking for supplements, peptides or training programmes.

Then connect the rest of the record. Sleep, HRV, resting heart rate, recovery, blood oxygen, training activity and body composition. VitaApp can compare this year with last, explain what may be driving a shift, and point to the next thing worth checking.

VitaApp starts with biological age, calculated from nine blood biomarkers using the PhenoAge algorithm. Upload a blood panel and it parses the results in around ten seconds, maps each marker against longevity-focused ranges, and tracks the change over time. Alex’s latest panel covered 47 biomarkers. His records go back to 2018.

Most people already have years of health data. Sleep from a wearable. Blood work from a lab. Training history from another app. At DeSci Berlin, @a_miloski showed VitaApp, @vitaDAO’s attempt to on putting all of it in one record only you can read. A private longevity co-pilot, unveiled 🧵

The move toward running your own health decisions is not theoretical. It already went mainstream. GLP-1s normalized the weekly injection in under a year. @bryan_johnson turned his own body into an open experiment. A whole generation is tracking, dosing, and self-optimizing without asking anyone for permission. The demand is already here. The open infrastructure to do it safely is what comes next.

Friday Motivation: Building the system where the future of science is autonomous.

Full talk here: piped.video/eZcHqbmIT1E?si=Ex1a…

The shift, in four lines: Data access becomes computational rights. Raw data becomes verifiable outputs. Ownership becomes accountable use. Extraction becomes constrained learning. She was clear this is not solved end to end, and not just about decentralizing storage. It is an invitation to build it. Learn from the data without taking the data. That may be the infrastructure DeSci needs next.

Her proposal flips the default. Today: give me the dataset. It moves to a repository, a researcher or AI model works on it, and the person loses control. Instead: define what someone is allowed to do with the data. Which method, for what purpose, and what result is allowed to leave. Bring the code to the data, not the data to the code. The method travels in. The raw data never leaves. Only a permitted, verifiable output comes out.

She challenged three comfortable assumptions: FAIR data is not automatically ethical. When the source is a person's body, access has to be opt-in, not assumed. Encryption is not governance. It does not decide who accesses, for what purpose, for how long, with what accountability. Decentralization does not erase power. Power just moves to the keys, the access paths, the cache, and the computation.

We think data ownership means who owns the dataset. But ownership hides inside the infrastructure. The real question is who controls the conditions of use: access, discovery, copying, computation. Health data makes this impossible to ignore, because both sides are right. Researchers genuinely need easier access. Patients genuinely need protection from extraction.

So her question is not how do we open more data. It is this: how can research learn from data without owning it, while the patient stays in control of who can ever access it. DeSci has built rails for funding, IP, and coordination. As AI agents get more capable, the bottleneck moves to a new layer. Who is allowed to compute over sensitive data. That is the next fight.

Here is the part that stayed with her. Even after the data finally moved, she could not find out who held a copy or who could still access it. The official channels took over a year. Then a private clinic just said: email us the documents, that is enough. Her health data now lives in places she cannot see. If she moves to Sweden, does her insurer get access and raise her premium? She still has no answer.

The story. She has vascular EDS, a genetic disease so rare that getting tested requires a paper trail proving you have the right to the test. She is Lithuanian, living in the Netherlands. Her sister, the first in the family diagnosed, is in Sweden. Her parents are in Lithuania. Three countries had to communicate before anyone could even be tested. Two years to gather the papers. A positive result. Another year for the data to reach her parents so they could be tested too.

It took @rakymnft two years just to get tested for a rare genetic condition that runs in her family. The result finally came through, and then it took another year to share it with her own parents. To this day, she has no idea who else has a copy of her DNA 🧵

Open, community-funded science is already producing real outcomes. Eli Lilly just bought Crossbridge Bio for $300M. The company grew out of one of the ecosystem's earliest community-funded research grants. Years in the making. Across the ecosystem so far: $2B+ in research IP traded, $50M+ into real science.

Two areas patients ask about most: brain and sleep. For cognition, Dihexa works through the HGF pathway and drove a large jump in synaptic connections in mice. People report feeling sharper and more engaged. Because it is strongly growth-promoting, he limits it to one to two months and screens carefully. For sleep, DSIP, a peptide found in high levels in mother's milk. Poor sleep tracks with higher cancer risk, so fixing sleep is rarely just about sleep.

GLP-1s are the most studied peptides, and the interesting part is what they do beyond weight loss. They interrupt the crave-and-reward cycle. In his practice that has helped patients quit smoking, drinking, and binge eating, because both the drive and the payoff fade. Early research even suggests GLP-1s alongside chemo may shrink tumors faster, likely by lowering the energy available to them. Roughly 1 in 10 Americans is now on some form of GLP-1.