koeng

31 ms·

361.
by koeng 6y ago
Whenever people talk about how GPT-3 can’t do a lot of things that humans can do, I always think back to the “Bitter Lesson”. I don’t want to believe that general AI will just come from a stupid amount of compute, but it might.
362.
by koeng 6y ago
Im a bit suspicious of the lactose experiment, mainly because there wasn’t hard data to back it up. Back then there was a spurt . AFAIK there are two people who have DIY gene engineered themselves and have actually gotten data to back it up
363.
by koeng 6y ago
Yep, pretty much. It’s a very interesting time - genetic therapies have extremely low marginal cost, but very high capital cost. Unlike software systems, they can mean life or death. Piracy of genetic therapies will be really interesting. S
364.
by koeng 6y ago
I used to be a mitochondrial engineer. This advance is fantastic, and not surprising. In 2006[0] they showed that you could get zinc fingers (similar to TALENs, which is what they use in paper) to site-specifically modify things in mitochon
365.
by koeng 6y ago
This lesson can be applied to synthetic biology right now, though it is still in its infant stages. At least a few of the original synthetic biologists are a bit disappointed in the rise of high-throughput testing for everything, instead of
366.
by koeng 6y ago
I would love to have something like that, but unfortunately that’s a really hard problem in itself, so I’d likely only start working on it once the need arises
367.
by koeng 6y ago
They won’t be. I definitely thought deeply about that aspect, but am choosing to not make them non-viable for the following reasons: 1. Payloads are in domesticated strains, which are less efficient than wild strains, which are already glob
368.
by koeng 6y ago
I’m trying to figure out how to ship DNA effectively. It’s really inefficient right now, so I’m genetically engineering bacterial spores to make it a lot better. Not sure if people will adopt, but at least it’ll be 10x better than what is c
369.
by koeng 6y ago
Until they show full genome sequencing of a novel strain under controlled conditions, I’d assume this is contamination or a crappy primer set. If you’re running a bunch of PCRs, especially of the same thing, there is a good chance of enviro
370.
by koeng 6y ago
Reordering of oligos is expensive, which makes people not reproduce. Sanger sequencing is pretty expensive ($5 sure, but over 5kb on 3 different constructs becomes pricey). PCR machines also ain’t that cheap (when you’re a noob and don’t kn
371.
by koeng 6y ago
NEB doesn’t ship residential, and there are many parts of the world that you can’t buy $20 oligos from. I like doing pretty complicated builds, and so I’d go bankrupt pretty much immediately because I don’t just build one thing.
372.
by koeng 6y ago
For doing mutagenesis on plasmids, you don't even need gibson. Overlapped primers do some weird circularization during PCR, so for point mutations you don't even need gibson mix. I'm personally a big fan of synthesis + Golden
373.
by koeng 6y ago
Really amazing resource that I’ve used before when pulling FP sequences. I really wish there were more bio databases like it! Simple, clean, great
374.
by koeng 6y ago
Remember that CRISPR is just an editing technique, and we have the technical capabilities to synthesize genomes. To get to cellular automata, I think there needs to be much more focus on building rather than tweaking. Though tweaking is ver
375.
by koeng 6y ago
Is beaker going to support IPFS?
376.
by koeng 6y ago
Could you give a couple examples of attacks that you thought was just theoretical but turned out to be very practical? Very curious about this topic.
377.
by koeng 6y ago
I found deploying databases with Dokku to be really intuitive. CockroachDB is great, but still a lot more steps than dokku postgres:create <db>. The whole certificates thing is quite confusing. Otherwise, k3s on-prem is great
378.
by koeng 6y ago
Would love to. Feel free to email me at koeng101<at>gmail.
379.
by koeng 6y ago
mdbook! By the folks making the Rust docs. I love their formatting.
380.
by koeng 6y ago
The dat website is at dat://aaca379867bff648f454337f36a65c8239f2437538f2a4e0b4b5eb389ea0caff You can visit with the beaker browser, or share it through dat so it won't ever go down. You can also visit it at http://
381.
by koeng 6y ago
Multiplexing. I use barcoded primers to amplify the sample, then pool and sequence
382.
by koeng 6y ago
Thanks for those links! I knew it was only a matter of time There are lots of folks working on LAMP in the DIYbio community. The kinda cool thing is that you can just have a colormetric read-out, so you don't even need Nanopore sequenc
383.
by koeng 6y ago
The key is to do barcoding at the PCR amplification step. That way, you can get away with barcoding hundreds of samples in a single tube. Really the prep screws it any other way. Is ONT coming out with solid state? Got a reference?
384.
by koeng 6y ago
Generally, yes absolutely. I’ve been doing a project called “NanoSavSeq” (Nanopore Saliva Sequencing) in my free time. It’s published on dat right now since the raw files for Nanopore are really big (got too big for hashbase). There is one
385.
by koeng 6y ago
Oxford nanopore sequencing. If a few problems can be figured out (mainly around machine learning and protein design), then it will beat every other biological detection, diagnosis, and sequencing method by a massive amount (no 10x, but more
386.
by koeng 6y ago
Current methods can pretty easily identify the virus from just your spit (in fact, that's a protocol that is pretty widely used for different applications). The biggest problem is that normally you need to concentrate the virus in orde
387.
by koeng 6y ago
You’re making a massive assumption there - that the IgG and IgM tests tell you if the person has been infected in the past. To which my experience says: no, it doesn’t. Both false positive and negative rates are very high in comparison to p
388.
by koeng 6y ago
C. elegans is a good example of this. They have males and hermaphordites, but interestingly enough, the ratios of those two from lab and wild strains differed by quite a bit. Turns out, in a stable lab environment, hermaphrodites were more
389.
by koeng 6y ago
> Imagine we had a biology forum that was a bunch of biologists that barely knew what a computer was discussing the latest media hype around AI. Would be pretty bad. Yea I’m around biotech forums quite a bit and it is just as bad from a
390.
by koeng 6y ago
Eukaryotes don’t really evolve very much on the protein level - their complexity allows them to mostly evolve on the regulatory level (while bacteria often still are on the protein level). Practically, this means that getting proteins from

More ›