Smash Notes

Back to OpenAI: When AI Agents Use Computers Faster Than You.

What Did OpenAI's Astra Change for Computer Use?

Ari Weinstein says Astra moved computer use forward at the edges even though computer-use models were already strong before Astra. It is a little faster on average because it uses fewer thinking tokens to do the same task. It is better at creative work, and it is more reliable around the edges.

In the launch video, someone talks to the model while it draws in Microsoft Paint, and Astra is way better at that sort of work. Weinstein thinks the improvement probably transfers to other creative tasks. He trusts it a little more. A small difference that is hard to perceive still matters between something being 99 percent reliable and 97 percent reliable, and Astra sits a little higher in a way that feels concrete.

OpenAI adds that Astra is also its most aligned model. The system card shows it is much better at resisting prompt injection and much better at avoiding unintended actions. For enterprise work on sensitive things, whether procurement, talking to vendors, or talking to customers, people are going to feel a lot safer using it.

Grokipedia·Microsoft Paint

Transcript

OpenAI: When AI Agents Use Computers Faster Than You ·

a16z

More from this episode

  1. Is OpenAI Computer Use Faster Than a Person?OpenAI says its computer use is now faster than the average person at using a computer.
  2. What Should Companies Automate With AI Agents?Ari Weinstein says companies should automate tedious workflows first, and the boring work is the opportunity.
  3. Why Will Crafted Software Still Matter With AI Agents?James Sun says crafted software will still matter alongside AI agents, because a product feels good when the person who designed it understands the experience…
  4. What Edge Do Startups Still Have Against OpenAI?James Sun says the moat startups still have against OpenAI is speed and pain.
  5. Why Wasn't OpenAI's Atlas Agent a Big Selling Point?James Sun says the agent inside OpenAI's Atlas was not a big selling point, because most people treated Atlas like a normal browser.

More answers like this

Related episodes

Plays from this note through the rest of the episode.

Weekly highlights

One letter a week.

The notes worth hearing, with the minute they were said.