Interesting – ChatGPT-4o has a special `ace_tools` module in its code execution environment, with convenience functions for displaying tables, interactive charts, etc., including calling out from the (sandboxed) code execution environment:
4
219
Excited to share some fun data structure work we did @OneCodex and recently open sourced: github.com/onecodex/rust-bfi…
Also on the front page of Hacker News for a hot minute... 🤞 news.ycombinator.com/news
Nick Greenfield retweeted
Under-appreciated leaderboard in my opinion - function calling (aka tool usage) is one of the most interesting applications of LLMs, plus it should be independent of how much "knowledge" is baked into the models
📊Delighted to welcome Command-R-Plus, Llama-3, and and Gemini-Pro-1.5 into the Berkeley Function Calling Leaderboard. Check out how they stack up across different categories, P95 latency, and costs at gorilla.cs.berkeley.edu/lead…
Congratulations to @cohere, @AIatMeta, and @GoogleDeepMind for pushing the Function Calling/Tool Calling boundaries even further 🚀
Nick Greenfield retweeted
Vedanta Biosciences Presents Progress from Two Defined Bacterial Consortia, VE303 for the Prevention of C. difficile and VE707 for the Prevention of Gram-Negative Infections, at IDWeek 2023 - businesswire.com/news/home/2…
Thank you for drawing attention to this @moseskagan. Have just donated as well. 🇮🇱💔
Friends:
Just made a $10,000 donation to American Friends of Magen David Adom, which provides emergency health services in Israel.
Have pasted the link below, in case you would like to join me.
Replies are off, to avoid turning this into a food fight among people I respect.
215
Nick Greenfield retweeted
I was recently invited by @Cancers_MDPI to write a microbiome-focused review. Very pleased to share our summary and recommendations on microbiome-based research in esophageal cancer.
pubmed.ncbi.nlm.nih.gov/3750…
#microbiome
#cancer
#metagenomics
#immunotherapy
Nick Greenfield retweeted
Blue blood and horseshoe crabs(Limulus): Limulus has existed millions of years, is cold-blooded and has a primitive immune system with 1 blood cell (ameobocyte), but why is their blood blue and why is it so valuable? (Pavid, K 2020 photo)
Nick Greenfield retweeted
Generative models and P vs. NP: A clickbaity thread 🧵
An important point that seems missing (as far as I've seen) in the debate over LLM "correctness" / "usefulness": generative models are most useful precisely in situations where creation is hard, but verification is easy. 1/N
Nick Greenfield retweeted
🧬Making DNA: The Quest for Synthetic Life
We're in the midst of a revolution in DNA synthesis.
But biology's grand plans — synthetic life, a solarpunk future — are hindered by major tech flaws & laughable biosecurity.
Read the 6,800 word essay.🔻
cell.substack.com/p/dna
Nick Greenfield retweeted
Accelerated computing is the future of genomics.
bsiranosian.com/bioinformati…
The volumes of data generated by modern seq experiments defies attempts to store and process it with traditional methods.
In this 3 part series, I'll cover basics, practical considerations, and algorithms
1/
Excited to share that I and a wonderful group of fmr colleagues (incl @heyaudy @DrShawnSolo @lynchde @clausmithdna on Twitter) have spun @OneCodex back out of Invitae! A wild ride, but grateful that I get to tag along with such a talented, dedicated group of ppl! 🙌 ❤️
We're thrilled to share that One Codex has been spun out as an independent company once again! 🎉🎊 Our team is as energized as ever, and we can’t wait to show you all the great new features we've been working on. Read more on our blog: buff.ly/3Sg8Ojh
@clausmithdna will be the GM, and I’ll be keeping my day job @Ginkgo Biosecurity, but excited to dip a toe back in the space and see @OneCodex continue to ship awesome software!
Nick Greenfield retweeted
The next wave of industry defining biotechs isn’t going to come from just having a platform or narrow uses of AI, but rather from full stack companies that *deeply integrate* new molecular technologies + hardware + software — in a similar way to how the iPhone was created. – at San Francisco, CA
Nick Greenfield retweeted
We need to start sequencing wastewater at scale. PCR is not going to cut it for wastewater based epidemiology to provide monitoring of current and future SARS-CoV-2 variants, independently from clinical samples.
Nick Greenfield retweeted
A bunch of genome scientists on the red carpet at the #TIME100 gala! @khmiga @timp0 @mike_schatz @roneilllab @EichlerLab … and my favorite celebrity, @kkariko! 🧬
Nick Greenfield retweeted
I've spent several years building software tools in genomics.
While the work couldn't be more exciting, there are unique challenges and tensions in the world of scientific software.
How does software in the life sciences actually work (and not work)?
newscience.org/how-software-…
Nick Greenfield retweeted
Our group @VedantaBio is hiring a Data Scientist to lead the analysis of clinical and pre-clinical microbiome studies. Support microbiome-based therapies in rCDI, IBD, cancer among other indications. Please RT. vedantabio.com/careers?gh_ji…
In times of crisis, it's important that we all look around and see how we can help. While this pales against the efforts of #HealthCareWorkers and dx labs, we are offering free analysis support to anyone sequencing COVID-19 samples.
Please DM or email if we can help in any way.
We're making an automated "one click" analysis of SARS-CoV-2 (COVID-19) genomics data available for free on One Codex: onecodex.com/blog/2020/03/16…
We plan to work closely with collaborators in dx and public health. Please DM or email [email protected] if we can help.
Nick Greenfield retweeted
It’s 2020. It’s a failure of imagination and will that we don’t prospectively sequence every American that gets sick from anything after this is over. It would cost <10’s of billions/year (<$1k/person & 10 MM tests/year). A small price to pay to be more aware of future pandemics.




























