At the HumanX conference, everyone was talking about Claude

At the HumanX AI conference in San Francisco this week, thousands of technologists gathered at the city’s Moscone Center to discuss how agent AI will change business. Agents that automate business and coding tasks are starting to be introduced across industries, primarily through chatbots for businesses and consumers.

Naturally, we wanted to know which chatbots were the most popular, and there was always one name we heard most often: It was Claude.

Anthropic was a hot topic at many of the panels held throughout the week, but also among the vendors we spoke to while browsing the convention room floor. A chatbot you’ve never heard of? Chat GPT. One vendor I spoke to emphasized that while he feels that ChatGPT and OpenAI are going downhill, or “declined” as the internet likes to say, he and his team use Claude heavily.

These days, that doesn’t seem like a particularly unique take. Indeed, despite a recent $122 billion funding round and an upcoming IPO, it’s not clear what will resolve the perception that OpenAI has lost its footing, or at least is becoming increasingly uncertain about what its next steps will be.

Part of the problem may be the perception that the company lacks focus. Last month, OpenAI decided to abandon a number of long-simmering side quests (including AI video generator Sora and difficult plans to launch a “sexy” version of ChatGPT) and instead focus on its business and coding services. On the other hand, a number of events have created a degree of negative buzz around the company, including a recent New Yorker article questioning whether the company’s CEO, Sam Altman, can be trusted. The company hasn’t won any friends by working with the Trump administration, nor has it won a decision to bring advertising to ChatGPT.

In one of the HumanX discussions, Sierra co-founder and CEO Brett Taylor (who is also OpenAI’s board chairman) defended Altman when asked by Alex Heath about the New Yorker profile. “I think Sam is one of the most visible leaders and executives in the world,” Taylor said. “If you’re going to look for people to detract from him, you’re going to find them, and they’re going to be very vocal about it,” he said, adding, “I think Sam is a great guy. I think he’s an outstanding leader in AI, and I really believe in his character as someone who’s worked with him.”

Controversy and turmoil can make OpenAI seem less strategic and more reactive, as if it were simply reacting to events rather than shaping them. That said, OpenAI and Anthropic are evenly matched in terms of name recognition and revenue. At least it looks that way. Some data suggests that Anthropic is catching up among business users. The Wall Street Journal recently analyzed the financials of both companies and found that they are “the fastest-growing companies in the history of technology.” In that sense, perhaps OpenAI’s “fall” means that it is no longer the undisputed champion. Competition exists, which is normal in most industries.

Rather, it remains clear that OpenAI is determined to do what it takes to maintain its edge. This week, the company announced a new $100 subscription tier for ChatGPT that significantly increases access to its coding tool, Codex. This move is apparently aimed at encouraging broader use of the tool while weaning users away from Claude Code.

In a HumanX discussion with Bloomberg reporter Rachel Metz, Srinivas Narayanan, OpenAI CTO for B2B applications, pointed out how rapidly the technology landscape is changing.

“We’re in this amazing moment in technology where every month, sometimes every day, we’re all looking forward to something new,” Narayanan said. Citing agent coding as an example, he added, “We knew AI would impact software engineering. People have been using assisted coding for the past year, but even in just the past few months, the entire field has changed.”

Other applications of AI (such as creative uses) haven’t really taken off yet, so agent outcomes may be a big focus of the technology community right now. Still, the amount of work that companies have started offloading to new small automated helpers is somewhat surprising, and as Narayanan noted in his remarks, it’s all happened in a relatively short period of time. Even in such an unpredictable environment, the future remains wide open.