Ginkgo Bioworks CEO Wants Biology to Manufacture Physical Goods | Hacker News Reader

Ginkgo Bioworks CEO Wants Biology to Manufacture Physical Goods

bloomberg.com

50 pointsdpflan16 comments

People sometimes ask why there's so much discussion of Lisp on HN. It's because so many times when you scratch the paint off a trend, a language, or a prominent person in computer science you find Lisp right below the surface.

One of the founders of Ginkgo, Tom Knight, was one of the inventors of the Lisp Machine.

I knew the name ringed a bell, thanks for filling in :)

> We have a partnership with a large Canadian company called the Cronos Group, a $100 million deal. We’re basically programming cells to produce cannabinoids. It’s very similar to what we did back in the fragrance industry. You get cannabinoids—THC, CBD, and so on—by extracting the oil from the flower of the cannabis plant, and then that goes into things like vaping and edible products. The process for doing all that growing is extremely expensive and requires big greenhouses. What we’re doing is reading the DNA of the cannabis plant and finding the part that encodes the cannabinoids. We then move it into brewer’s yeast, brew it up, and instead of beer, you get CBD.

I noticed he mentions CBD but not other parts of the plant, they also mention CBD on their homepage. Hopefully they don't spend too much resources on the quack science stuff which a lot of CBD stuff is being marketed with (outside of epilepsy).

The other stuff they are working on sounds truly fascinating. The amount of money going into it is pretty crazy, I guess they've made a proven model with mint oil / brewers yeast and are now scaling it up like a tech startup. Or rather like YC for bio startups but more hands-on with IP ownership and providing resources.

These guys have been on a major PR push for a little while. The business model and technology sound fantastic, and the potential applications really are numerous.

I find the notion that ‘anything’ can be commercially produced by synbio dubious (think gold or uranium) but the range of applications in the organic space seem vast.

Does anybody have the relevant background to reliably comment on the cost efficiency/R&D timelines/scalability of these kinds of projects?

I work in bioinformatics, not syn-bio, but I have friends working in syn-bio.

I first wanted to go into syn-bio during my undergrad days (about 13 years ago?) since then all the exciting things were happening, logical gates, fine-grain control of E. coli etc.

When I look at what my friends are working on nothing much has changed? We can have simple AND gates in plants now, not just bacteria, but that's where it already stops. Anything more complex still doesn't work.

A lot of progress has been made in DNA synthesis, creation of new DNA instructions, 'printing' DNA?, we can write entire chromosomes now.

Delineating 'synthetic biology' from good old biotechnology is hard though :) A lot of the progress I see that gets chalked up to syn-bio feels more like regular biotechnology: getting known genes together into bacteria to produce chemicals. We've been doing that for a long time in insulin, washing powder, all kinds of medicines, all on industrial scale, it's where your insulin comes from.

To me syn-bio means to create entirely novel genes/pathways, not just re-using what nature has to offer, there was lots of fun progress here (artificial genetic codes, for example), but nothing that has made it out of the lab.

So Gingko's success happens in (to me!!!!) good old microbiology, taking genes that already exist and putting them into bacteria: 'Our first customers were in the fragrance industry. You get mint oil from mint leaves. We would take the genes from mint by reading the DNA code of the mint plant, find the part of it that encodes the mint flavor, print out the code from the mint plant, redesign it a little bit, move it into brewer’s yeast—like you use to make beer. ' That's not different from insulin production, and we've been doing that (well!) for thirty years or so. Same goes for Gingko's CBD production, we know a lot about that pathway already, and moving it into bacteria for easier production isn't impossible, I'd say feasible within the next 5 to 10 years.

Now making truly artificial synthetic biology, new genes, new pathways, logical switches, I'd say that's another 40-50 years until yo use that on industrial scale.

>Gingko's success happens in (to me!!!!) good old microbiology, taking genes that already exist and putting them into bacteria

My intuition suggests the same... although it’s based on high school AP bio plus conversations/reading.

Any thoughts on cost efficiency? Is there some Gibbs free energy or other principle that would prevent it from ever being commercially feasible?

Regular biotech is already very cost-efficient, it's just a matter of putting the gene in, not breaking anything, and then optimising the hell out of the protocol (optimising temperature, pressure, nutrient mix etc. in your bioreactors).

For synthetic biology, I have no idea why this is not commercially feasible. We probably oversimplify the cell's system too much, so that introducing a multi-level computer breaks things we don't know about? But I'm purely speculating.

There are zero profitable syn bio companies in the world because even simple projects are too slow and costly.

Ginkgo doesn’t value it’s employees, and doesn’t have secret sauce. I think the next company, that Ginkgo diaspora found or flee to, will be the one that starts making these claims real.

Similar vein between Amyris/Zymergen

You are correct. I was on an IGEM project when I was in undergrad, didn’t do too well but was amazed by the potential.

After I finished grad school, interviewed at 3 Syn Bio start-ups. Did not get offers from any of them, but all 3 failed within 1.5 years of me applying (blessing in disguise).

I find the technology and ramifications fascinating, but we are at least 20 years out from anything meaningful in the space.

I wonder if you might expand on your assertion that Ginkgo doesn't value its employees -- I have a friend that joined Ginkgo not too long ago, and I hope things work out for them.

It resembles academia. The company is very penny wise and pound foolish, toxic behaviors aren’t managed, the (5!) founders are basically dead weight, and hours are expected to bleed into evenings and weekends. There’s an inversion of resources- the more work you have, the less you will be supported by the company. The people who flourish tend to have academic backgrounds and this perpetuates the cycle. Good luck to your friend!

Having spent years working at a startup, you can confidently divide any and all claims made in the article by a factor of at least 2, if not 10. Remember that you're reading the most generous interpretation of what _could_ _potentially_ happen at some point in the future, and aren't reading about the failures and things that did not pan out. It's basically concentrated confirmation bias maximally generously extrapolated into the future. Think Elon Musk and "full self driving" for a calibrated reference. TL;DR: when a startup CEO or founder says something, don't believe a single word until you can actually verify the claims.

https://web.archive.org/web/20191109035053/https://www.bloom...

Does anyone know if Ginko Bioworks has a replicable process? Or is each synthetic organism built ad-hoc? Whats their special sauce / unfair advantage over others, like universities?

Weren’t DNA computers that could store petabytes in a thimble-full of material supposed to come out a decade ago? Biology is not easy harness, even if you “want” it...

Reply on news.ycombinator.com