Seth Howes

should have done this a long time ago

I expect that, if it isn't already, the bottleneck in most scientific fields will no longer be idea generation or data interpretation. It will be experimental throughput. From Eric Kauderer-Abrams, Head of Life Sciences at @Anthropic : “We couldn’t figure it out… It took us 3 months, ultimately, and lots of people working day and night in the lab to fix the problem. I posed this problem to Claude and, just in one minute, Claude actually one-shotted the answer."

@iou___man soon™️

@SethSHowes Do a better autopipetter next. Max volumetric accuracy

1/ Lab automation is insanely mispriced. Today I visited a London-based biomaterials startup running large-scale bacterial screens. They use a colony picker that costs $95,000

@techno0ptimist @PrimordiaGrants @NikoMcCarty Will take a look tysm

@SethSHowes Sounds like a potential project for @PrimordiaGrants ! Also check out the new grants announced by @NikoMcCarty to lower costs and increase speed in science.

@ghartoularos Why so?

@SethSHowes Do a flow sorter next

@benghunt @josiezayner it's just a flesh wound

@SethSHowes @josiezayner that's pretty long!

Props to @salonium for a super interesting article on the history of vaccine design in the latest edition of @WorksInProgMag worksinprogress.co/issue/the-gold…

In 1907 a neuroscientist cut a piece of spinal cord from a frog embryo, dropped it into a blob of frog plasma, sealed it with wax… …and accidentally invented a foundational technique in cell biology. On the discovery of the hanging drop method by Ross Harrison: “But scientists still couldn’t cultivate animal cells. They clumped together on a plate, ran out of nutrients, and quickly died. And as viruses are obligate parasites, requiring living cells to replicate, they couldn’t be cultivated this way at all. The next step forward came from neuroscientist Ross Harrison in 1907, who developed the hanging drop method. He took a piece of spinal cord from a frog embryo and put it in a drop of its plasma, placed it over a concave slide, and sealed it with wax - creating a drop that hung suspended and stayed moist and rounded while the plasma provided nutrients and structure. The nerve fibers stayed alive, growing for weeks.” Just run more weird experiments.

@benghunt @josiezayner go long or go home

@SethSHowes @josiezayner you could look at the stock prices of oxford nanopore, ginkgo, twist bio over the last year and see a curve

@CalcNeuro We’ve built these ourselves too. Similar OOM cost saving. So about 150 idiot index for plate readers (even worse than colony pickers)

@SethSHowes yeah, we had the same with a fluorescent plate readers: when we needed one we built it for about 2K instead of 100-150K. Now we have 5.

@jrkelly 💪💪💪📉

@SethSHowes get at it ! 🙌🙌🙌

@__hand_banana @josiezayner nod to @owl_posting for discovering this nugget of insanity and for writing a great post on current state of play for automated labs - owlposting.com/p/heuristics-f…

@__hand_banana @josiezayner Counter to this, apparently operating Hamilton liquid handlers is so non-trivial that they offer a 3-day course that costs $3.5k to teach you how to use them. (not a joke) - hamiltoncompany.com/services

@SimiStern 💪

@SethSHowes bullish, this is awesome

@CooperVeit3 why buy an expensive robot when you have phd hands amirite?

@SethSHowes Lots of companies to juicero

@_benray @ProfTomEllis oof these don’t retain their value on the used market given they cost upwards of $100k new

@SethSHowes @ProfTomEllis ebay.com/itm/1875460596…

@Bengineer_Cook 💪

@SethSHowes working on this too, let's talk!

@_benray @ProfTomEllis equally true for a liquid handler for which there is much greater demand. you can build one of these for $500 but the cheapest option on the market is an opentron which will set you back $20k

markets fail to serve biology because there are likely fewer than 1k people that would purchase a colony picker today at any price, even though the world would be better off having a large surplus of automation-friendly cheap scientific instruments the only goal of an instrument salesperson learns is to grift 100k off your startup

@noah_vandal Totally agree. Now super easy to generate these designs as well if you define them in code and sprinkle in your coding agent of choice - github.com/gumyr/build123d

@SethSHowes I’ve been thinking the same thing for 2+ years. So much of this could be a 3d printed assembly with a respberry pi

@josiezayner I wonder what the shape of the demand curve is for this hardware ie what latent demand is realised when you drop the price by 100x across all of these pieces of equipment?

@SethSHowes There is alot of hardware like this in Bio I think the problem is usually that the market is so small and sales channels so limited its hard to find a good reason to make some of this stuff