Bookmarks

observations on lab automation & biodata I see four archetypes of business resolving... 1. 'hardware' companies e.g., @trilobio robotics companies which exist to offer large scale deployments into pharma companies and biotechs. the product is the hardware. sustainable, but not growing at VC-backed pace - few of these left given they’ve all now pivoted in reaction to the demand for biodata. 2. 'output data' companies e.g,. @adaptyvbio infrastructure that allows rapid verification of biological hypotheses, giving agents the ability to generate ground truth on-demand. the experimental output data is the product. most sustainable over the long run - these will be the companies that allow agents to autonomously explore scientific search space. Strong long-term demand given how little we currently understand about biology and much more data scientific AIs will need to truly generalise across domains. 3. ‘process data’ companies - e.g., @transfyrai how we teach claude / GPT to be great lab scientists. the product here is data which trains agents *how* to do science - what the process looks like, how to troubleshoot, defining ‘good’ and ‘bad’ practice such that agents can one day perform or instruct wet-lab science themselves. limited sustainability - given the correct input data, models will quickly saturate. Marginal utility of this kind of data will fall off rapidly after some time. if the ‘process’ companies are constructed in the right way they will pivot seamlessly into ‘output’ companies (or indeed do both at the same time) 4. 'data dropshipping' companies - e.g., too many people right now micro1-esq plays for acquiring (typically pre-existing) datasets on demand. you'll be able to make a quick buck here if you have differentiated access to specific datasets or populations. least sustainable - given that many of these plays are based on monetising a limited set of data, not on the ability to generate more data over time.

Non-obvious traits of the best people I’ve ever worked with: 1) Short-term impatient, long-term patient 2) Don’t gossip 3) Impeccable with their word 4) Push me to think bigger 5) Manage their emotions extremely well 6) Deep sense of self-belief 7) Default to optimism 8) Balance earned confidence with deep humility 9) Play very intentional games & use their unique cheat codes to win 10) Bias to radical candor over ruinous empathy 11) Always managing health paranoia & OCD 12) Take "going down the rabbit hole" to the extreme

I still think a very unexplored area is drug repurposing, and it might be the right time to start a company here - This seems to be a priority for the FDA and they’ve put out a request for comments earlier this year for how they can incentivize development here. - ARPA-H is dedicating $76M to Every Cure to test new candidates. - AI tools are making it easier to connect disparate datasets and papers to find promising repurposing candidates which would normally cost a lot of money/human capital to sift. Many of these might be connections a human researcher would never think of making. TxGNN is a great example of this, making serendipitous discovery more concrete. - Simulation environments to test promising candidates are getting better with some of the newer virtual cell models like Arc Institute’s State.

good brands exist at the intersection of contradictions their audience feels caught between ex: - stripe: rigorous and romantic - patagonia: chill and ambitious - linear: obsessive and serene - gohar world: surrealist and functional - nike: rebellious and disciplined good

One of the most underrated side effects of betting on yourself is how spiritually wealthy you're forced to become. You spend all day navigating the uncertainty, trying to make sense of the discomfort Sometimes all you have is the 10 minutes of journaling before you close your eyes at night, the three sentences you repeatedly recite in your head, the random acts of kindness you started to gift to others because you need all the karma you can get You look forward to those intimate moments you get to have with yourself. The self belief you've learned how to generate feels so good it becomes a healthy addiction you can no longer live without. No matter how much you want to throw in the towel at times you just can't give up. To go back would mean you'd have to sacrifice that relationship with God

If you want to sell more in B2B, be less formal. Has worked for me. Stop sending claude slop decks. Stop sending polished emails. Break their frame. Can't tell you how many people have said something along the lines of "really refreshing that you guys did actual analysis and not an ai slop deck" Literally a basic ass word doc of screenshots and rambling notes + what this would cost to do...10-20k/mo deals from this consistently.

Excitement is the more practical synonym for happiness, and it is precisely what you should strive to chase. It is the cure-all.

"you need a new domain for every 2-3 cold email inboxes." we run up to 60 off one domain using subdomains. we tested it across ~30,000 accounts first. these are the rules that held up: the root: • one clean .com per tree, bought from Porkbun or Namecheap • it gets forwarded through Cloudflare • it NEVER sends, not even one campaign • it's only there so the branches have something to hang off the branches: • each subdomain looks like send.yourdomain. com • 20 branches per root is the hard ceiling • every branch holds 2 to 3 official Google or Microsoft inboxes • SPF, DKIM, DMARC and MX are set on each branch before it sends • a flagged branch gets cut off and the rest keep going the sending: • each inbox sends 10 cold emails a day • each inbox also sends 20 warm-up emails a day • warm-up and campaigns start together and run side by side • i build to about half the ceiling, so roughly 10 branches per root where to be careful: • somebody sends from the root and one burn takes every branch down with it • the root is a .info, which is fine on its own but too shaky to hang subdomains off • a root gets filled all the way to 20 branches, with no margin left • you only run ~30 inboxes across the whole business. this earns its keep at hundreds or thousands of accounts a few of our customers send from ~10,000 inboxes this way.

At Works In Progress we are always looking for new writers with smart, fresh takes the world needs to hear. Could you be one of them? We publish longreads on tech, policy, economics, architecture, culture, and all sorts of things. If you write for us your work will be printed in our beautiful magazine, read by many of the world's most interesting people. We also pay well. Here are some examples of the sort of articles we would love to commission: - Why isn't there more semiconductor manufacturing in Silicon Valley? - Fifteen years ago, CRISPR was an incredibly promising new medical technology. What has happened to all the CRISPR drugs? - Japan was once the stock example of a low-fertility country, but it has staved off further decline as other East Asian countries have plummeted past it. How? - Poland has had exceptional post-communism economic growth, and is on track to overtake Britain in wealth per capita in 2030. What caused the Polish economic miracle? If you have an idea for something you think is the sort of thing we would publish - or the sort of thing we *should* be publishing - please get in touch, or send a pitch to pitches@worksinprogress.co.

the original lock-in…

As a woman who generally gets approached in annoying ways, HOW you do the approaching matters and is what makes the experience acceptable vs. annoying or creepy. Examples that would work on me: "Nice purse." (When she says thanks tell her you have the same one at home with a

sales gets a lot easier when you stop looking for advanced shit and just become brutally consistent at the basics: - prospect daily, even when your pipeline looks good - prepare for every call like it is a high-stakes conversation - ask a clarifying question before responding to anything unclear - disqualify fast when the fit, urgency, or authority is not there - leave every call with a clear next step, owner, and date - follow up on the buyer’s decision process, not just for an update - review your calls, especially the ones you thought went well - update your pipeline based on buyer evidence, not hope - ask for feedback even when you do not want to hear it - do what you promised without needing to be reminded the fundamentals are boring. that’s why almost nobody gets fucking elite at them.

Millen Rastogi walked away from a career at Deutsche Bank to take over his mother’s struggling home care business in New York & Connecticut. A paper-based, compliance-heavy business that was barely surviving. He breaks down how he fixed the initial company and then executed 5 add-on acquisitions in 24 months, often at extremely attractive multiples, using a scrappy capital stack and heavy seller financing. All with the goal to (maybe) sell the platform at an 8-12x multiple to middle-market private equity. Show notes: 0:00 Millen Rastogi 1:05 From Deutsche Bank to home care 4:30 Walking into a mess 13:26 Funding the early years 19:54 Growth numbers + COVID hit 23:02 Margins and revenue 25:50 Building the platform 27:17 First acquisition 30:56 Seller financing playbook 36:35 The “silver tsunami” sellers 41:50 Second acquisition 51:48 70% seller-financed deal example 1:09:09 Learning M&A with AI + X

The most valuable advice we got during YC was to explain our company using verbs instead of nouns. Early on, I walked into a pitch and did the complete opposite: “We’re reinventing the cloud for AI” No one had any clue what that meant. I could (very painfully) see their eyes

free things to do in new york after work that aren't a bar: • moma, fridays 5:30 to 8:30 (free for ny state residents, reserve ahead) • the whitney, fridays 5 to 10pm, and always free if you're 25 or under • the morgan library, fridays 5 to 8pm (reserve ahead) • the frick, first fridays 5:30 to 9pm, next one's oct 2 • museum of the city of new york, all day wednesdays, open till 9 • free shows at the rubenstein atrium at lincoln center, no tickets needed • juilliard students playing free rush hour sets at broadway & 66th • the stavros niarchos library on fifth, open till 9pm mon to thu • simons foundation science talks on fifth ave, oct 7, 14 and 21 • madison square park, open till midnight through october • brookfield place winter garden, open 24/7 and if it's past midnight and you still don't want to go home, you know where we are ;)

My favorite use cases for Muse / Instinct so far: 1. Submit FOIA requests to request data from the US government 2. Creating spend-limited Privacy cards to spend on subscriptions without having them recur 3. End to end filed an entire visa form for a country 4. Responded to a coordination mail for a wedding by finding the flight and hotel details 5. Look for reservations for restaurants or concerts when they open and purchase them immediately A lot of the web was designed with dark patterns: increase friction to prevent enough humans from doing something, and now those walls are completely broken. At this point, I feel like I’m squarely limited by creativity and understanding what is possible.

Only 53 shows on IMDb with 9.0+ and 25k+ reviews. Only 10 premiered in the last 5 years. I watched one of them, a 1 season 16 episode Korean show called When Life Gives You Tangerines, and it’s the most beautiful family drama tear-jerker ever. One of the best shows ever.

I talked to a founder who was demoralized because all the ideas he could think of could be easily duplicated. I told him to ignore that worry; nearly all startup ideas are like that at first. Their value is the other ideas they lead to.

i keep a running doc called “things i’m not doing anymore” and look at it monthly really simple way to live a happier & more productive life

i could never have articulated this but damn have i felt it

There is no formula to winning anything worthwhile. Sometimes you end up a godforsaken city you never intended on ending up in and decide to play a practical joke on a bunch of nerds in a library and one of those nerds becomes the love of your life. Sometimes a random DM cascades into a whole new life. Sometimes you try and try and try and nothing pans out and then you stub your toe on something you never knew you needed. You have to confidently surf the insane randomness and mischief of this universe. The love of the game is all there is.

1. If you had a heart attack and had to work two hours per day, what would you do? Not five hours, not four hours, not three—two hours. It’s not where I want you to ultimately be, but it’s a start. Besides, I can hear your brain bubbling already: That’s ridiculous. Impossible! I know, I know. If I told you that you could survive for months, functioning quite well, on four hours of sleep per night, would you believe me? Probably not. Notwithstanding, millions of new mothers do it all the time. This exercise is not optional. The doctor has warned you, after triple-bypass surgery, that if you don’t cut down your work to two hours per day for the first three months post-op, you will die. How would you do it? 2. If you had a second heart attack and had to work two hours per week, what would you do? 3. If you had a gun to your head and had to stop doing ⅘ of different time-consuming activities, what would you remove? Simplicity requires ruthlessness. If you had to stop ⅘ of time-consuming activities—e-mail, phone calls, conversations, paperwork, meetings, advertising, customers, suppliers, products, services, etc.—what would you eliminate to keep the negative effect on income to a minimum? Used even once per month, this question alone can keep you sane and on track.

Someone once told me the following: The reason to succeed isn't only for money, power, or some other form of leverage. It's because if you're the best at what you do, you'll spend time with others who are the best at what they do. And that's a full life. It's why chefs spend time with musicians, athletes with technologists, investors with artists, and more.

Leonardo da Vinci's greatest paragraph