AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
SOMETHING BIG IS COMING AT OPENAI DEVDAY.
DevDay is next Tuesday.
So… what are they building?
Managed autonomous agents?
Agents with their own computers and environments?
Persistent memory?
Teams of agents working together?
Codex that can build, test AND deploy entire products?
Or something nobody has leaked yet?
OpenAI already has the Agents API, hosted sandboxes, tools, skills and frontier-level reasoning.
Now connect the dots.
What if DevDay is where all of this starts becoming ONE system?
You don’t ask AI how to do the work.
You give it the goal.
It plans.
It codes.
It researches.
It uses tools.
It works for hours.
It delivers the result.
From AI assistant → AI worker.
And if Astra allowed OpenAI itself to build this “most ambitious sprint” faster than ever…
imagine what happens when millions of developers get the same leverage.
Tuesday will be a very interesting day.
SOMETHING BIG IS COMING AT OPENAI DEVDAY.
DevDay is next Tuesday.
So… what are they building?
Managed autonomous agents?
Agents with their own computers and environments?
Persistent memory?
Teams of agents working together?
Codex that can build, test AND deploy entire products?
Or something nobody has leaked yet?
OpenAI already has the Agents API, hosted sandboxes, tools, skills and frontier-level reasoning.
Now connect the dots.
What if DevDay is where all of this starts becoming ONE system?
You don’t ask AI how to do the work.
You give it the goal.
It plans.
It codes.
It researches.
It uses tools.
It works for hours.
It delivers the result.
From AI assistant → AI worker.
And if Astra allowed OpenAI itself to build this “most ambitious sprint” faster than ever…
imagine what happens when millions of developers get the same leverage.
Tuesday will be a very interesting day.
SOMETHING BIG IS COMING AT OPENAI DEVDAY.
DevDay is next Tuesday.
So… what are they building?
Managed autonomous agents?
Agents with their own computers and environments?
Persistent memory?
Teams of agents working together?
Codex that can build, test AND deploy entire products?
Or something nobody has leaked yet?
OpenAI already has the Agents API, hosted sandboxes, tools, skills and frontier-level reasoning.
Now connect the dots.
What if DevDay is where all of this starts becoming ONE system?
You don’t ask AI how to do the work.
You give it the goal.
It plans.
It codes.
It researches.
It uses tools.
It works for hours.
It delivers the result.
From AI assistant → AI worker.
And if Astra allowed OpenAI itself to build this “most ambitious sprint” faster than ever…
imagine what happens when millions of developers get the same leverage.
Tuesday will be a very interesting day.
SOMETHING BIG IS COMING AT OPENAI DEVDAY.
DevDay is next Tuesday.
So… what are they building?
Managed autonomous agents?
Agents with their own computers and environments?
Persistent memory?
Teams of agents working together?
Codex that can build, test AND deploy entire products?
Or something nobody has leaked yet?
OpenAI already has the Agents API, hosted sandboxes, tools, skills and frontier-level reasoning.
Now connect the dots.
What if DevDay is where all of this starts becoming ONE system?
You don’t ask AI how to do the work.
You give it the goal.
It plans.
It codes.
It researches.
It uses tools.
It works for hours.
It delivers the result.
From AI assistant → AI worker.
And if Astra allowed OpenAI itself to build this “most ambitious sprint” faster than ever…
imagine what happens when millions of developers get the same leverage.
Tuesday will be a very interesting day.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.
AI IS STARTING TO DISCOVER THINGS WE DIDN’T KNOW EXISTED.
Claude just identified a previously unknown enzyme system hidden inside bacteriophage DNA.
Next to it: a strange repeating DNA structure with similarities to CRISPR.
Nobody knows what it does yet.
And THAT is exactly why this is so interesting.
This wasn’t Claude summarizing a paper.
The system wasn’t in the literature.
Claude searched biological data, generated a hypothesis - and Anthropic’s new molecular biology lab is now testing AI-generated discoveries in the physical world.
Think about where this goes.
Today: one unknown enzyme system.
Tomorrow: thousands of AI scientists searching genomes, proteins, molecules and chemical spaces simultaneously.
Every discovery becomes data for the next experiment.
Every experiment improves the next hypothesis.
AI → experiment → discovery → AI → experiment → discovery.
Run that loop 24/7. At machine speed. Across millions of hypotheses.
New medicines?
New materials?
New biological mechanisms?
Longevity?
What happens when scientific discovery itself becomes exponential?
Maybe superintelligence won’t first reveal itself by passing another benchmark.
Maybe it reveals itself by discovering things humanity never knew were there.