conv.

All stories
AIActive today · day 3

Anthropic's Claude flags a new enzyme system; CRISPR scientists split on its significance

Anthropic's biology lab says nearly 1,000 Claude agents found a CRISPR-like system in phage DNA, but outside researchers call it routine genome mining.

Part of a larger narrative

OpenAI Trade Secrets Theft

3 stories · since Sep 14 · newest 4h ago — OpenAI faces mounting legal and regulatory pressure over allegations that its engineers misappropriated proprietary technology and trade secrets from major…

  1. Anthropic's Claude flags a new enzyme system; CRISPR scientists split on its significanceyou are here
  2. Anthropic launches Claude Opus 5.5, cutting price 40% while lifting coding performance
  3. Anthropic taps Accenture for AI safety evaluations ahead of November IPO

What to know

  • Anthropic's first wet-lab result: Claude autonomously flagged an uncharacterized enzyme system tied to a CRISPR-like DNA repeat array, but its biological function remains unknown.
  • The company frames this as proof AI can speed up fundamental biology; independent researchers are split between excitement (Feng Zhang, Stanley Qi) and skepticism that it's routine genome mining (Kevin Blake).
  • The announcement lands as Anthropic expands into drug discovery and courts scientists and investors ahead of a possible public listing.
  • Online discussion is largely skeptical, questioning both the anthropomorphic framing of Claude's 'discovery' and whether the result is genuinely novel versus repackaged data-mining for PR.

The dispute Whether the finding is a genuine autonomous scientific discovery or a repackaged, PR-timed result from routine data mining. · positions read across 85 posts and comments

most voices

The announcement oversells routine genome mining as a breakthrough, and the narrowing search process may be discarding real novelty.

  • “So Going from 200k RTs to 3500 candidates to 20 reports to 1 real hit very narrowing path basically it seems like the real risk ; the things thrown out of this filter as an outlair might be an…”

    falcon_tech · Hacker News ↗
some voices

Anthropic's language subtly anthropomorphizes Claude to imply agency or sentience it doesn't have.

  • “the subtle, almost deceptive or manipulative, ways these companies choose words to nudge along the narrative that their LLM systems are conscious/sentient/persons/etc”

    1659447091 · Hacker News ↗
some voices

AI-driven science is a genuinely valuable use of the technology, but the company's profit motive undercuts the good news.

  • “This article highlights a rare positive impact of AI. AI has the potential to revolutionize scientific/medical research. That is the only justifiable use. Unfortunately, it doesn’t line the pockets of the tech bros. They foist all the…”

    just-jorj.bsky.social · Bluesky ↗
some voices

The discussion should avoid extreme framings of 'meaningless' versus 'singularity' and stick to what the discovery process actually shows.

  • “I wish we could discuss this in a way that didn't immediately devolve into people shouting up or shouting down that this is either meaningless or singularity.”

    nonameiguess · Hacker News ↗

“What stands out is its ability to recognize an unusual biological pattern that was difficult to detect before, and to pursue it comprehensively as a research question”

Stanley Qi, Stanford bioengineering professor · Al Jazeera ↗ · Sep 23

Dario AmodeiDario Amodei Anthropic CEOAnthropic AI company running the biology labFeng Zhang CRISPR pioneer, MIT/Broad Institute professorStanley Qi Stanford bioengineering professorKevin Blake Microbiologist, Washington University School of Medicine

Anthropic's Claude flags a new enzyme system; CRISPR scientists split on its significance
anthropic.com

How it unfolded 6 developments, newest first · click a bar or a number to jump articlespostscomments

Peak 32 pieces in one hour at Sep 23, 2 PM; 158 pieces over 3 days (49 articles · 45 posts · 64 comments) Sep 22, 5 AM — 1 piece · 1 article — Newswires 1Sep 22, 6 AM — quietSep 22, 7 AM — quietSep 22, 8 AM — quietSep 22, 9 AM — quietSep 22, 10 AM — quietSep 22, 11 AM — 1 piece · 1 article — Newswires 1Sep 22, 12 PM — quietSep 22, 1 PM — quietSep 22, 2 PM — quietSep 22, 3 PM — quietSep 22, 4 PM — quietSep 22, 5 PM — quietSep 22, 6 PM — quietSep 22, 7 PM — quietSep 22, 8 PM — quietSep 22, 9 PM — quietSep 22, 10 PM — quietSep 22, 11 PM — quietSep 23, 12 AM — quietSep 23, 1 AM — quietSep 23, 2 AM — quietSep 23, 3 AM — quietSep 23, 4 AM — quietSep 23, 5 AM — quietSep 23, 6 AM — quietSep 23, 7 AM — quietSep 23, 8 AM — quietSep 23, 9 AM — quietSep 23, 10 AM — quietSep 23, 11 AM — quietSep 23, 12 PM — quietSep 23, 1 PM — 3 pieces · 2 articles · 1 post — Mastodon 2, Newswires 1Sep 23, 2 PM — 32 pieces · 19 articles · 9 posts · 4 comments — Newswires 19, X 5, Hacker News 5, +2 moreSep 23, 3 PM — 16 pieces · 1 article · 2 posts · 13 comments — Hacker News 13, Mastodon 2, Google News 1Sep 23, 4 PM — 8 pieces · 1 article · 3 posts · 4 comments — Hacker News 4, Mastodon 3, Newswires 1Sep 23, 5 PM — 4 pieces · 4 comments — Hacker News 4Sep 23, 6 PM — 20 pieces · 6 articles · 5 posts · 9 comments — Hacker News 9, Google News 5, Mastodon 4, +2 moreSep 23, 7 PM — 5 pieces · 1 post · 4 comments — Hacker News 4, Mastodon 1Sep 23, 8 PM — 5 pieces · 2 posts · 3 comments — Hacker News 3, Bluesky 1, Mastodon 1Sep 23, 9 PM — 4 pieces · 3 articles · 1 comment — Google News 3, Hacker News 1Sep 23, 10 PM — 3 pieces · 2 posts · 1 comment — Mastodon 2, Hacker News 1Sep 23, 11 PM — 3 pieces · 2 articles · 1 comment — Newswires 2, Hacker News 1Yesterday, 12 AM — 6 pieces · 4 articles · 2 posts — Newswires 3, Mastodon 3Yesterday, 1 AM — quietYesterday, 2 AM — 11 pieces · 3 posts · 8 comments — Hacker News 8, Mastodon 2, Bluesky 1Yesterday, 3 AM — 6 pieces · 4 posts · 2 comments — Mastodon 3, Hacker News 2, Bluesky 1Yesterday, 4 AM — 4 pieces · 2 articles · 2 comments — Hacker News 2, Google News 1, Newswires 1Yesterday, 5 AM — 3 pieces · 1 post · 2 comments — Hacker News 2, Mastodon 1Yesterday, 6 AM — 2 pieces · 1 article · 1 comment — Hacker News 1, Newswires 1Yesterday, 7 AM — 2 pieces · 1 post · 1 comment — Hacker News 1, Mastodon 1Yesterday, 8 AM — 5 pieces · 1 article · 3 posts · 1 comment — Mastodon 2, Hacker News 2, Newswires 1Yesterday, 9 AM — 2 pieces · 2 comments — Hacker News 2Yesterday, 10 AM — 2 pieces · 1 post · 1 comment — Hacker News 1, Reddit 1Yesterday, 11 AM — 3 pieces · 3 articles — Newswires 3Yesterday, 12 PM — 3 pieces · 1 article · 2 posts — Bluesky 1, Mastodon 1, Newswires 1Yesterday, 1 PM — quietYesterday, 2 PM — 1 piece · 1 post — Reddit 1Yesterday, 3 PM — quietYesterday, 4 PM — 2 pieces · 1 article · 1 post — Mastodon 1, Newswires 1Yesterday, 5 PM — quietYesterday, 6 PM — quietYesterday, 7 PM — quietYesterday, 8 PM — 1 piece · 1 post — Mastodon 1Yesterday, 9 PM — quietYesterday, 10 PM — quietYesterday, 11 PM — quietToday, 12 AM — quiet 12–345–6
Sep 23yesterdaynow · 1:05 AM ET
  1. 6

    Online audience ties the claim to Anthropic's own AI-safety warnings

    Social commenters noted the biology announcement arrives just after Anthropic and other AI CEOs said their models had become capable and dangerous enough that the industry must slow down, prompting questions about consistency and motive.

    “the revelation that Anthropic has been conducting biology experiments comes right after AI CEOs, including Amodei, publicly admitted that models have become so capable, and so potentially dangerous, that the industry must slow down...”
    — j11h11.bsky.social
    • Carl_Zimmer@mastodon.social

      Anthropic's new biology lab released a technical report on using AI to discover new enzymes. Experts I talked to say "discover" is doing a lot of work here. Story:

      Carl_Zimmer@mastodon.socialMastodon4h ago15▲view on Mastodon ↗
    2 more of the top 3 · 3 posts in this stretch
    • just-jorj.bsky.social

      techcrunch.com/2026/09/23/a... This article highlights a rare positive impact of AI. AI has the potential to revolutionize scientific/medical research. That is the only justifiable use. Unfortunately, it doesn’t line the pockets of the tech bros. They foist all the useless slop on us for profit

      just-jorj.bsky.socialBluesky12h ago6▲view on Bluesky ↗
    • persagen@mastodon.social

      # LLM - Claude discovers a novel enzyme system | Anthropic https://www. anthropic.com/news/claude-disc overs-novel-enzyme-system Claude discovers a novel enzyme system with CRISPR-like repeats https:// news.ycombinator.com/item?id=4 9820134 Impressive ... 🤔

      persagen@mastodon.socialMastodon12h agoview on Mastodon ↗
    all of them →
  2. 5

    CRISPR researchers call the finding routine genome mining

    Coverage highlighted that while Anthropic compared the system to CRISPR, some CRISPR-field scientists said such repeat sequences are common and undercatalogued, and that nothing indicates the discovery is a rival to CRISPR-the-technology or has practical application.

    1. first by Anadolu Agency, 18h ago · also Quartz, The Decoder

      2 more headlines
    • Thats a facinating step in AI usage into different aspects. Just curious How will this process guarding against the failure mode where agents say " this is unusual " , and the judgement is itself biased towards known family . Its easy to flag something as intresting when it partialyl resembles an already kown system (CRISPR- like repeats , in this…

      falcon_techHacker News15h agoview on Hacker News ↗
    2 more of the top 3 · 6 posts in this stretch
    • Considering what happened on the Navier–Stokes event of OpenAI recently, I keep wondering how much of these discoveries are actuallya.) novel isolated achievements of an AI, orb.) the result of continuous focused in-house training with data involuntarily contributed by thousands of researchers using the LLM, aiming to make a press-release to boost…

      rickdeckardHacker News15h agoview on Hacker News ↗
    • > For now, this is mostly a story about how AI can be used to parse existing data to discover new biology (which is fantastic!).I'd like to expand that: in my view, this is also a story of how agentic AI systems can come up with bioinformatics strategies to discover novel features. One would think such a task would be the ideal domain of the…

      mfldHacker News16h agoview on Hacker News ↗
    all of them →
  3. 4

    Amodei and outside scientists react, split on significance

    Dario Amodei posted that the system could represent a new gene-editing mechanism and said he is 'rooting for' frontier labs to pursue biology, while Stanford's Stanley Qi called the pattern-recognition 'incredibly exciting' and microbiologist Kevin Blake called the CRISPR comparison overstated.

    “I’m genuinely rooting for all of the frontier labs to seriously get into biological discovery, and I’m excited about what comes out of it.”
    — Dario Amodei
    1. first by Gizmodo, 1d ago · also Indian Express, Reuters, Channel News Asia, Times of India, Al Jazeera, Hindustan Times

      5 more headlines
    • bitsontape@privacysafe.social

      🔬 The best bytes of # science & # tech across the # fediverse “Anthropic's biolab made a discovery it's comparing to Crispr theverge․com/ai-artificial-intelligence/999470/anthropic-biolab-claude-crispr ⌗AI ⌗Science ⌗Research” https://www. theverge.com/ai-artificial-int elligence/999470/anthropic-biolab-claude-crispr 🤖 via RSS feed. Not an…

      bitsontape@privacysafe.socialMastodon21h ago5▲view on Mastodon ↗
    2 more of the top 3 · 29 posts in this stretch
    • j11h11.bsky.social

      Good AI work? "Still, the revelation that Anthropic has been conducting biology experiments comes right after AI CEOs, including Amodei, publicly admitted that models have become so capable, and so potentially dangerous, that the industry must slow down..." #AI

      j11h11.bsky.socialBluesky22h ago2▲view on Bluesky ↗
    • I've been using Claude Science a lot and it is VERY good at finding patterns in the DNA around my binding sites - quite often it went 'you could put your primer here but that looks like an Alu repeat, so better not, the primer won't be specific' - it seems like the press release is one step above that pattern recognition? I.e., 'there's a…

      a_bonoboHacker News1d agoview on Hacker News ↗
    all of them →
  4. 3

    Anthropic details the scale of the automated search on X

    Anthropic's official account said nearly 950 Claude agents spent 21 hours and 210 million tokens searching DNA sequence data before one agent flagged the unusual repeat pattern next to the enzyme's gene.

    “Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages.”
    — Anthropic (@anthropicai)
    • JUST IN: Anthropic’s Claude has said it has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages

      @unusual_whalesX1d ago109▲view on X ↗
    2 more of the top 3 · 46 posts in this stretch
    • Why is everyone quick to point out how blogs/articles are "ai slop", but no one blinks an eye at the subtle, almost deceptive or manipulative, ways these companies choose words to nudge along the narrative that their LLM systems are conscious/sentient/persons/etc? The systems they are creating are impressive enough on its own merit. There is…

      1659447091Hacker News1d agoview on Hacker News ↗
    • cdarwin@c.im

      Last week, Anthropic confirmed that it was operating a wet biology lab in the Bay Area where it uses its AI models to run physical experiments. This week, the AI giant announced that this lab has already made what it believes is a big discovery: a new enzyme “system” that has certain “properties reminiscent of CRISPR,” as Anthropic describes it…

      cdarwin@c.imMastodon1d agoview on Mastodon ↗
    all of them →
  5. 2

    Anthropic announces Claude discovered a CRISPR-like enzyme system

    Anthropic published results from its new molecular biology lab describing an enzyme system, based on a reverse transcriptase in a jumbo phage, that sits beside a repeating DNA array reminiscent of CRISPR; the company says Claude worked with only high-level direction from scientists.

    “We gave Claude a prompt to search through a massive database of DNA sequences for interesting new examples of RTs. Our involvement was limited to the initial prompt and the lab work…”
    — Feng Zhang
    1. 6 outlets first by Blockchain.News, 1d ago · also Moneycontrol, The Verge, Anthropic, HN Best, HN Frontpage · read ↗

    2. first by Breitbart, 12h ago

    2 more claims →
    • BREAKING: Anthropic’s new wet lab produces its first discovery, with nearly 1,000 Claude agents autonomously uncovering a novel enzyme system after 21 hours of work.

      @PolymarketX1d ago9.3k▲view on X ↗
  6. 1

    Anthropic reveals it built a wet lab for Claude-guided experiments

    Ahead of the discovery announcement, reporting surfaced that Anthropic had set up a physical biology lab where Claude guides robots through drug-related experiments, extending its life-sciences work beyond simulation.

    1. first by Phys.org, 20h ago · also Straits Times, The Economic Times

    2. first by The Decoder, 2d ago · also TechRepublic

      1 more headline
    1 more claim →

Also covered reported alongside — the timeline has no entry for these yet

  1. first by Seeking Alpha, 1d ago · also The Verge

    1 more headline

and 4 smaller pieces

What people are saying 12 voices from 4 sites · best of 85 · verbatim

Still unanswered
  • Is the enzyme system actually novel, or just an uncatalogued but common CRISPR-like sequence?
  • What safeguards exist against agents biasing toward flagging things that merely resemble known systems like CRISPR?
  • Why is Anthropic publicizing a discovery whose function is unknown, especially while also warning that its models are becoming dangerously capable?