nycdatasci

Opus 5.5 was a step function change. Really crushes on multi-hour coding compared with prior models.

Punctuation? Not our specialty.

Excerpt from the Project Defuse document (March 2018): “we will intensively sample bats at our field șites where we have identified high spillover risk SARSr-CoVs. We will sequence their spike proteins; reverse engineer them to conduct binding assays, and insert them into bat SARSr-CoV (WIV1, SHC014) backbones (these use bat-SARSr-CoV backbones, not SARS-CoV, and are exempt from dual-use and gain of function concerns) to infect humanized mice and assess capacity to cause SARS-like disease.“

March 24, 2018: “we will intensively sample bats at our field șites where we have identified high spillover risk SARSr-CoVs. We will sequence their spike proteins; reverse engineer them to conduct binding assays, and insert them into bat SARSr-CoV (WIV1, SHC014) backbones (these use bat-SARSr-CoV backbones, not SARS-CoV, and are exempt from dual-use and gain of function concerns) to infect humanized mice and assess capacity to cause SARS-like disease.”

Post from Mark Chen for those not on X:

  Two things to distinguish:
  Did any human or agent look at user data as part of the Navier Stokes effort? No.
  Do we use user feedback and de-identified data to improve ChatGPT and Codex in a holistic way? Yes. And so does every LLM company.