John Wilbanks
@wilbanks@wilbanks on Bluesky @johnwilbanks at https://t.co/esSpM3fecN
Similar User
@Sagebio
@petersuber
@atulbutte
@jasonpriem
@lindaavey
@kristiholmes
@jenny_molloy
@MarkHahnel
@openscience
@eperlste
@PLOS
@MikeTaylor
@hjoseph
@thePeerJ
@fredtrotter
Applications for Astera's 1st residency cohort just opened! This is a bonkers opportunity for ppl who want to build public goods to advance science & tech: salary for a year, budget for teams, ample compute resources, opportunities for follow-on $ -- please share w/your networks!
💫Applications are open for Astera’s first residency cohort💫 We are seeking creative, high-agency scientists, engineers, and entrepreneurs passionate about building open projects for public benefit 🧵
🥁🥁🥁🥁🥁 Check out the new open call for the @AsteraInstitute residency program! Are you a builder that wants to help shape the future for the benefit of all? Do you share our ethos of impact through openness? Do your ideas not quite fit existing structures? This on-site…
💫Applications are open for Astera’s first residency cohort💫 We are seeking creative, high-agency scientists, engineers, and entrepreneurs passionate about building open projects for public benefit 🧵
Seriously good opportunities in these residencies - please share widely!
💫Applications are open for Astera’s first residency cohort💫 We are seeking creative, high-agency scientists, engineers, and entrepreneurs passionate about building open projects for public benefit 🧵
💫Applications are open for Astera’s first residency cohort💫 We are seeking creative, high-agency scientists, engineers, and entrepreneurs passionate about building open projects for public benefit 🧵
Astera's Open Science Fair will be Thur, Nov 14, 6-8:30pm, in Berkeley: docs.google.com/forms/d/e/1FAI… (org by @jessicapolka & @AsteraInstitute )
In 2003 the hot thing in biology was genomics and protein folding was considered an intractable, unrewarding problem As we all herd on to the AI bandwagon, it is worth remembering that the exact opposite behavior is what led to real innovation pubmed.ncbi.nlm.nih.gov/14631033/
Big shout out to those researchers who made their data available in PDB over the years The Chemistry Nobel Prize winners today relied on that open academic data to train their models We wouldn't have these advances if researchers hadn't made 200,000 protein structures available
A theory of change for philanthropic investing asterainstitute.substack.com/p/philanthropi…
How can philanthropy power change in emerging technologies? Kyle McEneaney is joining Astera as our first Head of Strategic Investments to drive our approach in catalytic philanthropy for science and technology 🧵
Shoutout to all the teachers who pour their brilliance into us! I celebrated this #MacFellow recognition by visiting my first school, Angeles Mesa “Where Dreams Take Flight” and where lunch was my favorite subject.☺️
Meet Ruha Benjamin (@ruha9), a transdisciplinary scholar, writer, and 2024 #MacFellow exploring the ethical and social dimensions of innovation. Learn more about Ruha ⬇️ macfound.org/fellows/class-…
We join more than 100 leading air quality organizations and colleagues in committing to share AQ data freely and openly and support policies that let others do the same. Check out this community letter and join us: docs.google.com/forms/d/e/1FAI… #OwnYourAQData
I don’t think this is just idle navel gazing. I hear a lot of people talking about how LLMs shouldn’t lie or spread misinformation and I think in order to properly apply those ideas you need a strong theory about what facts and LLM, for lack of a better word, knows.
Does the truth of P matter for how I interpret this? What if I restart the session and it says “not P”? Does this imply anything about its belief that P is true? What if it generates both “P” and “not P” in the same session?
One sub-question is: if the LLM spits out a string “P” that I understand as a proposition P, how should I interpret that? Is the LLM implicitly saying “P is true”? Is it saying “I believe P is true”? Or is it just meaningless text that signifies nothing without my interpretation?
It’s pretty hard for me to wrap my mind around what it means for a proposition to be “believed” by an LLM, let alone how to determine whether that belief is “justified”.
I’m relatively comfortable intuitively saying something like ChatGPT seems to “know” that the sky is blue but I really have no idea what I mean by that. The old standby would be I’m asserting that it has justified true belief that the sky is blue. Can this be made to make sense?
Has anyone come up with a way to talk about whether an LLM “knows” a fact?
Data Products: A new specification is born The Data Product (DPROD) specification is a profile of the Data Catalog (DCAT) Vocabulary Candidate specification to be submitted to the OMG; on #Github H/T @Andrea_Gioia #Metadata #Standards #OpenSource github.com/EKGF/dprod
Welcome @wilbanks! Super excited to get this off the ground and put 💰 where our 👄 is when it comes to opening up science via data @AsteraInstitute
We believe the future is open. Open data coupled with the explosion of compute will accelerate what is possible with science. We’re thrilled that John Wilbanks @wilbanks is joining Astera as our first Head of Data to drive this vision.
Back on here for the big announcement, and will be around a bit for work! #Scitwitter is anyone still around?
United States Trends
- 1. #SmackDown 41,3 B posts
- 2. Jared McCain 9.723 posts
- 3. Khalid 23,8 B posts
- 4. Kendrick 815 B posts
- 5. Bayley 3.929 posts
- 6. #OPLive 1.534 posts
- 7. Bianca 16,8 B posts
- 8. MSNBC 272 B posts
- 9. #loveafterlockup 1.649 posts
- 10. #AskShadow 31,7 B posts
- 11. Kevin Owens 2.997 posts
- 12. Chelsea 50,9 B posts
- 13. Scott Bessent 23,7 B posts
- 14. Ben Simmons 2.773 posts
- 15. Creighton 2.741 posts
- 16. Shinsuke 2.085 posts
- 17. Chaz Lanier N/A
- 18. Cam Johnson 1.829 posts
- 19. #OPNation N/A
- 20. Raquel 13,6 B posts
Who to follow
-
Sage Bionetworks
@Sagebio -
Peter Suber (@[email protected])
@petersuber -
Atul Butte
@atulbutte -
Jason Priem
@jasonpriem -
Linda
@lindaavey -
Kristi Holmes
@kristiholmes -
Jenny Molloy
@jenny_molloy -
Mark Hahnel
@MarkHahnel -
Open Science
@openscience -
Ethan Perlstein bio/acc
@eperlste -
PLOS
@PLOS -
@[email protected] 🏴 🇬🇧 🇪🇺 🦣
@MikeTaylor -
Heather Joseph
@hjoseph -
PeerJ Publishing
@thePeerJ -
fredtrotter
@fredtrotter
Something went wrong.
Something went wrong.