6 min read

Jumper joins Anthropic; safety gets a roadmap

Jumper joins Anthropic; safety gets a roadmap
Nº 01 · The Lede Hacker News Field report

Jumper jumps to Anthropic

Jumper jumps to Anthropic
Fig. IHacker News · Filed 22 Jun 2026.

John Jumper joined Anthropic, the AlphaFold co-creator who shared the 2024 Chemistry Nobel leaving Google DeepMind for the safety-focused frontier lab behind Claude. The move pulls one of the most credentialed names in protein-structure AI out of the lab that built AlphaFold and into a generalist model shop with no public bio program — a signal that frontier-lab work on biology is consolidating around whoever can pay for the talent, not the labs with the longest bio track record. It also resets the recruiting reference: Anthropic just became a plausible destination for senior bio-AI researchers in a way it wasn't last week.

Read the source

DeepMind publishes control roadmap
Fig. IIX · Filed 22 Jun 2026.
Nº 02 X Field report

DeepMind publishes control roadmap

Google DeepMind released an AI Control Roadmap built on the premise that frontier models may not do what their operators intend — a framework for monitoring, containing, and intervening on misaligned behavior in deployed systems. Publishing the assumption out loud reframes the safety debate from alignment-by-training to control-by-design, and sets a reference document any lab deploying autonomous agents on sensitive biological or clinical data will be asked to compare against.

Read more
TxBench-PP scores preclinical pharmacology
Fig. IIIX · Filed 22 Jun 2026.
Nº 03 X Clinical AI · Evaluation

TxBench-PP scores preclinical pharmacology

TxBench-PP benchmarks models on small-molecule preclinical pharmacology with verifiable answers — the first focused slice of a broader therapeutics evaluation effort. It anchors a real reference point for the "our model is good at drug discovery" claim that vendors have made unfalsifiable for two years — a gap we flagged when TxBench-PP first appeared — turning preclinical pharm capability into something a buyer can actually score.

Read more
Also Filed · Five Briefs from the queue
Nº 04 Hacker News Computational biology

Genome 3D structure trips up AI

Quanta argues the human genome's physical 3D folding — loops, territories, contact domains — encodes regulatory logic that sequence-only language models miss. Pushes back on the assumption that scaling DNA language models on linear sequence will keep delivering, and reframes what a "foundation model for genomics" needs to ingest to keep improving.

Read
Nº 05 arXiv Agents · Infrastructure

Dual agents translate lab protocols

A two-agent system translates natural-language protocols into executable robotic-platform code, with one agent generating and a second verifying across models before execution. Moves protocol-to-robot translation from single-shot LLM output toward verified pipelines — the missing reliability layer for autonomous wet labs.

Read
Nº 06 bioRxiv Field report

Chain-aware PLM predicts antibody affinity

Antibody-antigen affinity prediction gets a protein language model that knows which chain it's reading — heavy, light, or antigen — instead of treating the complex as one sequence. Narrows the gap between general PLMs and antibody-specific tools, with affinity scoring inching toward something usable inside an active design loop.

Read
Nº 07 bioRxiv Field report

GENATATORs annotates genes ab initio

GENATATORs uses DNA language models to call genes from raw sequence without homology evidence, a long-standing gap for newly assembled or non-model genomes. Raises the floor for what's annotatable on day one of a new assembly.

Read
Nº 08 arXiv Clinical AI · Evaluation

MedRLM stacks clinical reasoning modalities

MedRLM combines long-context clinical reasoning, sensor-guided screening, and referral routing in one recursive multimodal system aimed at community-to-tertiary care handoffs. Adds another entrant to the crowded clinical-LLM field where deployment evidence, not architecture, is now the differentiator.

Read

Reply with your discoveries. A human reads them. Forward freely.

Agentic Discovery  ·  Nº 40  ·  22 Jun 2026

Editor's Note

Monday opens with a Nobel laureate switching jerseys and DeepMind admitting the AI-doesn't-cooperate scenario out loud.

 

Nº 01 · The Lede  —  Hacker News  —  Field report

Jumper jumps to Anthropic

Jumper jumps to Anthropic

Fig. I  Hacker News · Filed 22 Jun 2026.

John Jumper joined Anthropic, the AlphaFold co-creator who shared the 2024 Chemistry Nobel leaving Google DeepMind for the safety-focused frontier lab behind Claude. The move pulls one of the most credentialed names in protein-structure AI out of the lab that built AlphaFold and into a generalist model shop with no public bio program — a signal that frontier-lab work on biology is consolidating around whoever can pay for the talent, not the labs with the longest bio track record. It also resets the recruiting reference: Anthropic just became a plausible destination for senior bio-AI researchers in a way it wasn't last week.

Read the source →

Why it matters

The center of gravity for protein-AI talent is no longer pinned to the lab that started it — every frontier shop is now a credible home for biology work, and bio-specific labs lose their monopoly on the people who built the field.

 

Nº 02  —  X  —  Field report

DeepMind publishes control roadmap

Fig. II  X · Filed 22 Jun 2026.

DeepMind publishes control roadmap

Google DeepMind released an AI Control Roadmap built on the premise that frontier models may not do what their operators intend — a framework for monitoring, containing, and intervening on misaligned behavior in deployed systems. Publishing the assumption out loud reframes the safety debate from alignment-by-training to control-by-design, and sets a reference document any lab deploying autonomous agents on sensitive biological or clinical data will be asked to compare against.

Read more →

 

Nº 03  —  X  —  Clinical AI · Evaluation

TxBench-PP scores preclinical pharmacology

Fig. III  X · Filed 22 Jun 2026.

TxBench-PP scores preclinical pharmacology

TxBench-PP benchmarks models on small-molecule preclinical pharmacology with verifiable answers — the first focused slice of a broader therapeutics evaluation effort. It anchors a real reference point for the "our model is good at drug discovery" claim that vendors have made unfalsifiable for two years — a gap we flagged when TxBench-PP first appeared — turning preclinical pharm capability into something a buyer can actually score.

Read more →

 

Also Filed  ·  Five Briefs from the queue

Nº 04  —  Hacker News  —  Computational biology

Genome 3D structure trips up AI

Quanta argues the human genome's physical 3D folding — loops, territories, contact domains — encodes regulatory logic that sequence-only language models miss. Pushes back on the assumption that scaling DNA language models on linear sequence will keep delivering, and reframes what a "foundation model for genomics" needs to ingest to keep improving.

Read →

Nº 05  —  arXiv  —  Agents · Infrastructure

Dual agents translate lab protocols

A two-agent system translates natural-language protocols into executable robotic-platform code, with one agent generating and a second verifying across models before execution. Moves protocol-to-robot translation from single-shot LLM output toward verified pipelines — the missing reliability layer for autonomous wet labs.

Read →

Nº 06  —  bioRxiv  —  Field report

Chain-aware PLM predicts antibody affinity

Antibody-antigen affinity prediction gets a protein language model that knows which chain it's reading — heavy, light, or antigen — instead of treating the complex as one sequence. Narrows the gap between general PLMs and antibody-specific tools, with affinity scoring inching toward something usable inside an active design loop.

Read →

Nº 07  —  bioRxiv  —  Field report

GENATATORs annotates genes ab initio

GENATATORs uses DNA language models to call genes from raw sequence without homology evidence, a long-standing gap for newly assembled or non-model genomes. Raises the floor for what's annotatable on day one of a new assembly.

Read →

Nº 08  —  arXiv  —  Clinical AI · Evaluation

MedRLM stacks clinical reasoning modalities

MedRLM combines long-context clinical reasoning, sensor-guided screening, and referral routing in one recursive multimodal system aimed at community-to-tertiary care handoffs. Adds another entrant to the crowded clinical-LLM field where deployment evidence, not architecture, is now the differentiator.

Read →

 

· · ·

Reply with your discoveries. A human reads them. Forward freely.