In partnership with
Around 950 Claude agents spent 21 hours searching DNA data.
One noticed a repeating pattern in the DNA of viruses that infect bacteria. Scientists are now investigating what it does. Anthropic
That is enough to make me interested.
AI helping researchers notice something worth testing is a story I want to follow, even while the answer is unfinished.
I am Alex. Welcome to ShortCu8 by Innov8.
Lets Dive Deep🐰🐰.
⭐Today's Shortcut
How the search worked
Anthropic's scientists gave Claude a research brief about reverse transcriptases, a family of enzymes that copy RNA into DNA. Think of the task as searching a huge biological catalogue for unusual entries worth a closer look.
The agents divided the work, checked results and opened follow-up investigations. The campaign involved 949 agent sessions over 21.5 hours.
While reading DNA beside one enzyme's gene, an agent noticed repeated sequences. It followed that clue and produced a report for scientists to review. Research paper
The interesting bit was beside the enzyme
Researchers had identified the underlying enzyme before. According to Anthropic, Claude noticed overlooked features around it: the repeat pattern and a nearby partner gene.
The team calls the system ART, short for array-associated reverse transcriptases. Early lab experiments showed that its repeat region produces short RNA molecules. The resemblance to CRISPR makes it worth investigating, but researchers still have to establish its function. Anthropic
Calling this a new gene-editing tool would get ahead of the evidence.
The team has shared the findings as a preprint. Human scientists carried out the lab experiments. Anthropic
They tried the search again
The team repeated the full search campaign ten times. Every rerun missed the repeat array. None inspected the stretch of DNA beside the relevant enzyme genes where the repeats were found.
This limits how reliably the search can reproduce its own success. It also leaves two separate jobs: understanding the biology, and improving the AI's ability to find promising leads consistently. Research paper
The ShortList
🛠️Cool Tools of the Week:
Qwen Intelligence: Personal intelligence agents by Alibaba for planning, mobile use and creative tasks.
Gemini Connected Apps: Google is adding several new apps to Gemini, including Airtable, Linear, Squarespace, Peloton and more.
YouTube Custom Feeds: Use AI to create dedicated tabs on YouTube for specific moods.
📩 Innathe Shortcu8 engane undarunnu 👇️?
We read every reply - just reply to this email and let us know how we can improve !
Appo adutha Shortcu8il kanaam bie…👋
If you read till here, you might find this interesting
#AD1
Some teams never seem to stop moving. They're on Attio, the agentic CRM.
Every customer signal is captured in one shared context layer, always current and compounding. Agents and workflows build pipeline, chase every buying signal, and move deals forward, an always-on revenue engine running alongside your team.
With Attio, you’ll get:
Leads automatically prioritised and routed to the right rep
Expansion and risk signals caught the moment they land
Follow-ups written in your voice, already there when you arrive
Teams like Parallel, Turbopuffer, and Wordsmith build on Attio. Are you one of them?
#AD2
DNSMOS gives you a score, not whether that data fits your model. Voices' CTO DJ Jalali breaks down the four-step framework the team uses to set model-specific quality thresholds instead.