Katy Shi is a researcher at OpenAI who studies Codex’s default behavior. She says while many people describe it as aloof, she believes that this personality can be changed. “dry bread,” Its less apologist style has been appreciated by many. “A lot of engineering work is about being able to take critical feedback without interpreting it as mean,” Shi says.
Codex is also being used by several major companies. “The fact that ChatGPT is synonymous with AI gives us a massive advantage in the B2B market,” You can find out more about this by clicking here. Fidji SimoOpenAI’s Director of Applications, Mr. “Companies want to use technologies their workers are already familiar with.” Simo explained that OpenAI’s plan to sell Codex relies heavily on bundling it with ChatGPT, and other OpenAI software.
Jeetu patel, Cisco’s chief product officer and president, has said that he told his employees to not worry about Codex’s cost, as they will need to become comfortable using the software. Employees often ask whether Codex is free. “they’re going to lose their job because they’re using these tools,” Patel says, “what we have to tell our people is no, but I guarantee you’ll lose your job if you don’t use them, because you won’t be relevant. So you’re going to be out.”
This morning, there is a panic AI coding agent discussion has gone beyond Silicon Valley. The Wall Street Journal credited Claude Code for causing a $1 trillion tech stock sell-off Investors feared software will soon be obsolete. IBM’s share price had its lowest day in over 25 years a couple of weeks after Anthropic said that Claude Code would be able to upgrade legacy COBOL-based systems. OpenAI worked hard to integrate its AI agent into the social conversation. They spent millions of dollars in a Super Bowl ad about Codex rather than ChatGPT.
Codex is not a topic of discussion at the Mission Bay Temple. OpenAI engineers told me that they no longer type code. The engineers spend most of their time talking to Codex. Sometimes they do this in congregation.
At headquarters, I sat in on a Codex hackathon—about 100 engineers crowded into a large room. The four-hour deadline was for everyone to make the best Codex demonstration. OpenAI’s senior leader stood in the middle of the room and spoke team names while twisting his laptop away. Representatives of the teams nervously approached a podium to give short speeches on their AI projects. Patagonia back packs were awarded to the winners.
Many projects have been created in Codex or designed to make engineers more efficient with Codex. One group developed a tool for converting Slack chats into weekly report. A second group created a guide for OpenAI’s internal services that was generated by AI in the style of Wikipedia. These demonstrations used to take weeks or days, but can now be created in just an afternoon.
Kevin Weil was standing in the hallway as I left. He is the new head of OpenAI for Science – the unit that builds AI products for scientists. He informed me that Codex had been working overnight on a few projects for him, and would review them the next morning. It’s a practice that Weil has adopted, along with hundreds of others. OpenAI has set a goal for 2026 to create an automatic intern who will do research in (what else?). AI.
