At the HumanX conference, everyone was talking about Claude

This week at the HumanX AI conference in San Francisco, thousands of techies descended on the Moscone Center, where the discussion focused on how agent-based AI is transforming business. Agents that automate business and development tasks have begun to be deployed across industries, primarily through enterprise and consumer chatbots.

Naturally, I wanted to know which chatbot was the most popular, and consistently I heard one name the most: Claude.

Anthropic gained recognition on many panels throughout the week, but it was also a topic of discussion with the salespeople I talked to while scanning the conference room floor. A chatbot I haven’t heard much about? ChatGPT. One vendor I spoke to told me that he and his team used Claude a lot while he felt ChatGPT and OpenAI had declined — or, as the internet likes to say, “dropped out.”

It doesn’t seem particularly special lately. Indeed, it is not clear what will cure this belief, despite its recentness $122 billion financing round and him upcoming IPOOpenAI has lost ground — or at least it seems increasingly uncertain what the next step will be.

Part of the problem may be the belief that the company lacks focus. Last month, OpenAI abandoned a number of long-simmering side hustles (including the Sora AI video generator and a troubled plan to launch a “sexy” version of ChatGPT), focusing instead on its business and coding services. In the meantime, a number of events took place, including: the latest article from the New Yorker the question of whether the company’s CEO, Sam Altman, is trustworthy or not has raised some concerns amount of negative noise around the company. The company’s work with the Trump administration also didn’t win it any friends, nor did its decision to place ads on ChatGPT.

During one HumanX discussion, Sierra co-founder and CEO Bret Taylor (who is also chairman of the board of OpenAI) defended Altman when asked by Alex Heath about the Recent Yorker profile. “I think Sam is one of the most visible leaders and managers in the world,” Taylor said. “If you want to find his opponents, you will find them, and they will be very vocal about it,” he said, adding: “I think Sam is extraordinary. I think he’s an extraordinary AI leader and I really trust his character as someone who has worked with him.”

Controversy and hesitancy can make OpenAI seem reactive rather than strategic, as if it is simply reacting to events rather than shaping them. That said, when it comes to position and revenue, OpenAI and Anthropic are neck and neck – or at least that’s what it looks like, and some data suggests that Anthropic catches up with business users. Wall Street Journal recently analyzed its finances, showing that both companies were “the fastest-growing companies in the history of technology.” In this sense, perhaps the “fall” of OpenAI simply means that it is no longer the undisputed champion. There is competition, which is normal in most industries.

In any case, it remains clear that OpenAI is determined to do everything in its power to remain dominant. This week, the company announced a novel $100 ChatGPT subscription tier, giving you significantly more access to its Codex coding tool. This move appears to be intended to encourage wider exploit of the tool and hopefully steer users away from Claude Code.

During a HumanX discussion with Bloomberg reporter Rachel Metz, CTO of B2B app OpenAI, Srinivas Narayanan noted how quickly the technology landscape is changing.

“We are in this incredible moment in technology where every month, and sometimes every day, we all look forward to something new,” Narayanan said. Pointing to agent-based coding as an example, he added: “We knew AI was going to have an impact on software engineering, people have been using agent-based coding for the last year, but even in the last few months the entire field has changed.”

Agent achievements may be the focus of the technology community right now because other applications of AI (imaginative applications, for example) have yet to prove their worth. Still, the amount of work that companies have started dumping on their novel little automated helpers is a bit surprising — and, as Narayanan noted in his remarks, it all happened in a relatively miniature period of time. In such an unpredictable environment, the future is still wide open.