Gunnar retweeted

MIT CSAIL@MIT_CSAIL30 Nov 2025
United States
United StatesConnected via United States App StoreAccount-level information, not a live location or per-post device.
All major neural networks, in one chart.
v/The Asimov Institute
Gunnar retweeted
Biotech has a narrative void.
Biotech has potential for much much more futuristic and positive outcomes than most other companies. But biotech founders largely miss the opportunity, don’t have priors to narrative building, or see it as a waste of time.
If you’re building a company in biotech, we all know that will take a long time. So to keep people engaged, interested, and inspired by what you’re doing, you need to continue to remind them about what you’re doing and why it will pull the future forward.
Not only this, but with biotech funding much diminished from ZIRP highs, it’s hard to raise money. Thus you’re competing with every company on science (given) but also narrative. I’ve seen even small launches, blog posts, and preprints change the trajectory of countless companies and I want you to benefit too.
And because of this companies have opportunities to paint their version of the future or let others capture the attention in their vertical and take the talent and funding.
Why narrative matters:
Eyeballs on your work means that you have a higher ability to raise money, hire the best people, and get in front of customers. The other side is that if you don’t take up narrative space other founders will. Nucleus is famous for this where there’s more scientifically legitimate groups such as Herasight and Orchid, but not controlling the narrative means that there’s room for Kian at Nucleus to entirely control story of IQ embryo selection. The Far-UVC space has a high mindshare, controversial company, Nukit, which has allegations of unsubstantiated evidence.
TLDR: if you don’t control the narrative, another company will.
What Narrative is:
- Distilled legibility of your vertical.
- How your company will pull the future forward.
- How the industry will reorient itself around you.
- World building to bring people to the future your company will create.
The best founders normally use all of these positionings while talking to different sets of investors, hires and customers and have a library of stories they add to as the company and its mindshare grow. Customers are normally compelled by distilled legibility (eg. what makes your protein designs better than X), whereas hires and investors will want more future building, value accrual, and world building positioning.
Put another way, narrative is a way to increase your perception-to-reality ratio to raise more money and give your company a fighting chance for scaled success.
Narrative makes Schelling points:
We (especially @mhdempsey) have written a lot about category creation, delineating different properties that can lead to category creation such as technological moats, margin profiles, TAM expansion, industry turnover, and new markets. Expanding on this is the breakaway companies that are inextricably linked to categories such as “Palantir in government tech, Anduril in defense tech, OpenAI in frontier AI, Ginkgo Bioworks in Synbio, Recursion in computational bio, Bitcoin in crypto, Physical Intelligence in robotics, and more”. Distilled legibility has led all of these companies to become Schelling Points.
How you get narrative capture:
Write, post, have launch parties around model releases. Specifically the ways we see founders getting mindshare is by:
- Releasing pre-prints. There are so many preprints by companies that it’s now become a standard of sorts.
- Blogging about progress - @newlimit does an excellent job here releasing bi-monthly updates. Escalante blog also had great asthetic and opinion that circulated fairly far.
- Writing the company’s masterplan. How success in one area begets new and bigger opportunities. This is now a trite example but Elon uses the vision piece to rebut the main critiques of the company, giving him more legitimacy for his goals. Brian Armstrong talked about how Coinbase will be worth more in the future than it is today.
- Releasing internal communications. A famous internal memo at Slack came out about selling the innovation not the product and thinking about who you want your customers to become. This aligns employees and shows the world that you are building something different. From an internal founder memo at slack. If you're selling to Pharma, who do you want your Pharma customers to become?
- Having launch parties to get persuadables onboard and understanding what you’re doing. The Boltz team is excellent at this and has launch presentations in multiple cities for every model release.
- Launch videos for companies are becoming increasingly common. @themoldcompany and @alterego_io are great versions of this. You can also share launch videos for your high-impact advisors like @chaidiscovery .
This is a call to action to please world build, please share your incredible progress, please share your master plans!
The world will thank you.
Gunnar retweeted
What if we could autocomplete DNA based on function?
Today in @Nature, we share semantic design—a strategy for function-guided design with genomic language models that leverages genomic context to create de novo genes with desired functions.🧵
nature.com/articles/s41586-0…
Gunnar retweeted

Hannu Rajaniemi@hannu13 Nov 2025
United States
United StatesConnected via United States App StoreAccount-level information, not a live location or per-post device.
Since 2016, I have been building @HelixNano , a clinical stage biotech (and still my main gig), with Nikolai Eroshenko. Recently, HelixNano teamed up with OpenAI to push AI bio's limits. To our surprise, we saw models invent genuinely new wet lab methods (publication soon). 2/19
Gunnar retweeted
How to undo an R-loop: A molecular machine called the Smc5/6 complex helps repair an intricate form of genetic damage.
elifesciences.org/digests/96…
Gunnar retweeted
Today we report that an engineered skin bacterium, swabbed gently on the head of a mouse, can unleash a potent antibody response against a pathogen. Could lead to topical vaccines that are applied in a cream. @DjenetBousbaine led the charge... @Nature 1/55
Gunnar retweeted
Two stages of evolution... @GrantKinsler @gsherloc @PetrovADmitri use #ExperimentalEvolution in yeast to show that while 1st-step mutants appear pleiotropic in nature, 2nd-step mutations show a shift towards modular adaptation #PLOSBiology plos.io/4glXoXJ
Gunnar retweeted
The #hidden impact of #bullying and #harassment in #biotechnology and biomedical engineering: cell.com/trends/biotechnolog…
@TrendsinBiotech @MSUMD #industry
Gunnar retweeted
'The group that once prided itself on avoiding a hive mind is now steeped in its own kind of piety. The rationalist community is often accused of being cultish, but what it may be becoming is something more anodyne and less edgy than a cult: a fandom'
asteriskmag.com/issues/08/ra…
Gunnar retweeted
Our review of the use and potential of #microbial, tree-of-life, and #holobiont #eDNA and #eRNA for #biominotoring and #ecosystem assessment is now Open Access!
cell.com/trends/microbiology…
@TrendsMicrobiol with @lauren_colours, @veragfonseca, Andy Briscoe, Jens Boenigk, Guy Woodward
Gunnar retweeted
Expert’s dilemma: An increase in a scientist’s abilities for creative interdisciplinary thinking is almost inevitably linked to a loss in expertise and thus credibility in their home field, reducing the acceptability of the work to their peers
genomebiology.biomedcentral.…
Gunnar retweeted
Researchers used two 300-million-year-old specimens of Arthropleura to study the head of this giant arthropod and found it has features that make it a hybrid between a millipede and a centipede. bit.ly/3Yj572w
Gunnar retweeted
Great paper exploring how agents can secretly collude/communicate, bypassing security: coordinating without explicit communication (Schelling points), using crypto operations & mathematical structures (couplings) to hide info in innocent-looking messages. arxiv.org/abs/2402.07510
Gunnar retweeted
Our perspective on Hsp90 as a global modifier of the genotype-phenotype-fitness map, along with its implications for evolution in nature and clinical settings, is now out in JMB. We review the existing literature, identify open questions, contextualize our recent findings on Hsp90, and present new analyses that expand the current understanding of its evolutionary impact. Congrats to @jaguilarrod and Chris Jakobson. Check it out here: sciencedirect.com/science/ar…
Gunnar retweeted
A study in @NatComputSci projects e-waste growth from generative AI to reach a total accumulation of 1.2-5 million metric tons during 2020-2030 if no waste reduction measures are implemented. go.nature.com/4e2RDg7
Gunnar retweeted
Starting today, Crepuscular will be donating $10 monthly to the American Civil Liberties Union. If we were lucky enough to have our income grow, the donation will grow in kind. We have other plans in the works to boost marginalized voices - stay tuned.
Gunnar retweeted
Friends! Gather round and let us tell you about Issue 6! It's got dimensions, witchery, spiders, spaceships and resistance. Basically everything you need.
Mummer cover at by Alice M @itsbabypears
Gunnar retweeted
AI-generated images threaten science as they can create convincing scientific data with ease — publishers and integrity specialists fear a torrent of faked science. ‣ nature.com/articles/d41586-0… @Nature
Gunnar retweeted
Weird microbes could help rewrite the origin of multicellular life newscientist.com/article/245…




























