Welcome to the forefront of conversational AI as we explore the fascinating world of AI chatbots in our dedicated blog series. Discover the latest advancements, applications, and strategies that propel the evolution of chatbot technology. From enhancing customer interactions to streamlining business processes, these articles delve into the innovative ways artificial intelligence is shaping the landscape of automated conversational agents. Whether you’re a business owner, developer, or simply intrigued by the future of interactive technology, join us on this journey to unravel the transformative power and endless possibilities of AI chatbots.
After chat models and AI agents, Altman bets the next phase will be proactive AI running constantly in the background. But many companies are still struggling with the basics: costs are spiraling, products are too fragmented, and most users don’t even know what to ask.
At an enterprise event organized by OpenAI, CEO Sam Altman laid out a three-phase thesis for AI product development. Chat models like ChatGPT were the first major product. Agent-based systems like Codex are the second. The third phase takes the automation further: “proactive AI” that runs constantly in the background.
“I bet what comes next will be this idea of constantly running proactive AI,” Altman said, adding that “if there’s one thing to get ready for as the next phase over the next year, this is the one I would pick.”
The current agent phase has been “the biggest category” so far, according to Altman, “done in response to customer demand.” Users are increasingly confused about when to use chat, Codex, or an API, and struggle to pull together the right context and combine all the available plug-ins.
OpenAI wants to push agents even further, laying the groundwork to “enable everything else we want to do,” Altman said. The company is working on a kind of super app that combines agentic features from Codex with ChatGPT and other tools.
But there’s a growing problem: costs. At the start of the year, costs weren’t on anyone’s radar. Now cost has become “a huge issue,” Altman said, calling it “probably the second biggest theme.” He pointed to Uber, which said it burned through its entire AI budget for the year in Q1 alone. OpenAI wants to counter this by squeezing more value out of its models for less money. “I think we’ll have a lot of ways we can help people get more value for less spend,” Altman said.
Altman also flagged a usage problem. “Most people” simply don’t know how to use AI systems well. They’re aware they’re not using the tech enough and aren’t getting full value from it, he said.
“It’s hard to learn to work in new ways,” Altman said. “The activation energy’s a little too high.” Even when OpenAI can demo impressive use cases, most customers don’t actually use AI that way.
Proactive AI is apparently OpenAI’s answer here, too. If people won’t learn to use AI, just automate around them. In Altman’s vision, users wouldn’t even need to understand what AI can do.
“Can you get OpenAI to just be running as an agent in the background all of the time, connected to all my company’s context? And just don’t even put it on me to try to have to understand what it can do. Be useful to me,” Altman said.
That shift will need major adjustments, he added. Companies will have to change how they deploy AI, redesign their security protocols, and rethink compute allocation. An AI running nonstop with access to an entire company’s context puts very different demands on data protection than a chat window answering one-off requests.
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive “AI Radar” frontier report six times a year, full archive access, and access to our comment section.
Stay in the loop on AI. Clear, useful, no fluff.
Follow The Decoder for AI news, background stories and expert analyses.
Stay in the loop on AI. Clear, useful, no fluff.