Author:Biteye core contributor Amelia

On February 15, 2026, OpenAI CEO Sam Altman announced on the X platform that Peter Steinberger, the founder of OpenClaw, officially joined the OpenAI team and is responsible for promoting the development of the next generation of personal AI agents. OpenClaw, the open source project that took the world by storm just a month after its birth, will be handed over to the foundation and remain open source and independent.

Founder Peter Steinberger is a senior Austrian developer. He founded PSPDFKit in 2011 and successfully exited it for about 100 million euros. He has 13 years of entrepreneurial experience. OpenClaw (formerly known as Clawdbot, Moltbot) is an open source AI assistant project launched in November 2025. It gained more than 150,000 GitHub Stars in two months, becoming one of the fastest growing open source projects in the history of GitHub.

Peter Steinberger wrote in a blog published on the same day:

"I have played the game of building a company, invested 13 years of my life, and learned a lot. I want to change the world, not build a big company. My next mission is to build an intelligent agent that even my mother can easily use."

This sentence is like a heavy hammer, hitting the heart of the entire technology industry.

What's even more interesting is that this "ClawFather" wrote this in his Discord server rules:

"No chatting about finance or crypto."

This triggered an embarrassing question for the Web3 industry: Why does the encryption world, which has decentralized genes and shouts AI+Web3 logic, fail to grow an agent as pure and powerful as OpenClaw?

XHunt ranks 559th in global influence. On February 2, the description of ClawFather was added to the profile.

Looking back at OpenClaw: it was not "planned" but "milled"

1.1 从农场少年到PDF之王

Peter Steinberger's growth trajectory shows a typical path of technical genius, but it is also full of anti-commercial idealism.

He was born in 1988 in a farming family in Lower Austria, Austria. From an early age, he had a passion for taking things apart and reassembling them—not for commercial purposes, but out of sheer intellectual curiosity. While a student at the Technical University of Vienna, he founded the school's first Mac/iOS development course and worked with Apple Austria as a team mentor.

In 2011, 23-year-old Steinberger founded PSPDFKit, a technology company focusing on the development of iOS PDF frameworks. The company went on to become one of the best-known PDF solutions on the iOS platform, with the framework used internally by Apple and running on more than 1 billion devices.

In 2014, PSPDFKit received its first round of external investment. By 2021, Steinberger sold a majority stake after Insight Partners invested. The deal is estimated to be worth around €100 million.

But success was not lost on him.

1.2 Burnout and Rebirth: Three Years After Leaving the Technology Circle

In his 13 years at PSPDFKit, Steinberger has experienced the typical entrepreneur burnout.

"I work most of the weekend. As CEO, you are the trash can. You have to deal with problems that no one else can solve."

After exiting in 2021, he left the technology industry entirely. For three whole years, he didn’t write a line of code, didn’t attend a single meeting, and didn’t send a single tweet.

What has he done in the past three years?

According to what he shared in the podcast, he returned to the Austrian countryside to relearn how to "live" — not as a CEO, but as a human being. He reads, hikes, spends time with his family, and thinks about what he really wants.

In early 2023, when he reopened his laptop, he found that his passion for technology was still there, but he had completely disenchanted with "entrepreneurship".

1.3 "Holy Grail Moment": The Birth of Claudbot

After returning to programming, Steinberger began exploring modern web tools and AI. The turning point came when working on a personal project using Anthropic’s Claude Code.

"I fed my messy personal project into Claude Code, had it write specifications, and then generate code based on those specifications. This was my 'holy shit, this is shocking' moment."

This experience led him to develop Claudbot - a deeply personal AI assistant that interacts with users via WhatsApp and controls everything from smart lights to GitHub repositories.

But Claudbot soon ran into trouble. Anthropic believes the name infringes on their trademark rights. Steinberger did not argue and simply changed the name to Moltbot (meaning to shed its skin, suggesting rebirth).

However, the name Moltbot has sparked new controversy - some have pointed out that it is too similar to the name of an adult content platform. So, the name was changed for the second time: OpenClaw.

The name has finally been decided. And its mascot - a lobster wearing a space helmet - has become a cultural symbol of the open source community.

1.4 Hunted by the Crypto Community

What is more terrifying than the name-changing war is the "swarm harassment" of the crypto community. In the interview, Steinberg recalled that experience and said frankly: "I almost had a nervous breakdown and even wanted to give up the project."

Various apps in the crypto community are trying to "tokenize" everything. They swarm into Discord like swarms and frantically @ him on Twitter, causing his notification flow to be completely disabled. What's even more exaggerated is that they will send him a bunch of hashes and "force me to collect the money." Steinberg said bluntly: "I really have no interest in that set of things - first, I am not financially lacking at all; second, I don't want to support that system, because it is the worst kind of online harassment I have ever experienced."

So in OpenClaw's Discord server rules, Steinberger wrote this:

"No chatting about finance or crypto."

This rule is not specific to a specific incident, but is written into the community convention as a basic principle.

This attitude stems from his understanding of the nature of the project: OpenClaw is a tool, a tool that makes people's lives more convenient, not a financial speculation target.

1.5 Reject Meta, reject OpenAI acquisition

Before joining OpenAI, Steinberger faced several choices.

According to TechCrunch, Meta attempted to acquire OpenClaw but was rejected by Steinberger. OpenAI also made an acquisition offer, which was also rejected.

The plan he ultimately chose was to join OpenAI but not sell OpenClaw.

This means:

• Peter Steinberger joins OpenAI

• OpenClaw is transferred to an independent foundation and remains open source

• OpenAI commits to sponsorship, but does not control

• He can continue to serve as an OpenClaw consultant

这种安排在硅谷极为罕见。 Usually, when a big company wants an open source project, they buy it outright—like Microsoft bought GitHub or IBM bought Red Hat.

But Steinberger insisted on keeping OpenClaw independent.

Why OpenAI?

2.1 "I have played the entrepreneurial game"

Explaining his choice, Steinberger said:

"I can definitely imagine how OpenClaw could become a huge company. But that doesn't excite me. I'm a builder at heart. I've played the game of building companies, put in 13 years of my life, and learned a lot."

Behind this sentence is the sobriety of a person who has experienced a complete entrepreneurial cycle.

He knows how to raise money, how to expand, and how to exit. He knows how to turn a technology project into a company valued at hundreds of millions of dollars. But he also learned what the cost was - burnout, loss of life, becoming a "trash can."

More importantly, he knows that building a big company and changing the world are two different things.

PSPDFKit is a successful company, but there are limits to how much it can change the world.它让 PDF 在 iOS 上更好用了，但这远远不够。

What OpenClaw has the potential to change is: the way humans interact with AI.

2.2 Only by working together can we go further

Steinberger mentioned in his blog that the more he communicated with the OpenAI team, the more he discovered that "both sides share the same vision."

What is this vision? It is to make AI truly serve people, to make technology simple enough that anyone can use it, and to combine the open source spirit with cutting-edge research.

Collaborating with OpenAI is "the fastest way to bring intelligence to everyone." This is not an ideological choice, but a pragmatic one. OpenAI has the most cutting-edge models, the most powerful computing power, and the broadest user base. If he really wants to "let mothers use intelligent agents", OpenAI is the most realistic path.

OpenAI promises to hand over OpenClaw to an independent foundation, promises to sponsor but not control, and promises to let him continue to invest time in maintaining the project. These commitments impress a real builder more than any financial terms.

Why Web3 failed to catch OpenClaw

3.1 The "original sin" of Web3: Tokens take precedence over products

There is an open secret in the Web3 industry: the primary goal of most projects is to issue coins, not to make products.

This is not a criticism, but a fact. In the financing logic of Web3, Token is the core. You need to first design the Token economic model, then raise funds, then develop products, and finally exit on the stock exchange.

This logic leads to a structural problem: the product exists to support the token price, not to solve user problems.

Let's see how OpenClaw would fare if it were a Web3 project:

Assumed OpenClaw Web3 version:

• August 2025: Publish a white paper announcing the creation of a "decentralized AI assistant"

• September 2025: Received US$5 million in VC financing, valued at US$100 million

• October 2025: Release testnet and airdrop early users

• November 2025: Listed on a certain alpha, market makers pulled the market, the currency price soared 10 times, FDV reached 1 billion US dollars

• December 2025: The community begins to discuss "when will the stock be available"

• January 2026: The team was accused of insider trading, and the currency price plummeted

• February 2026: The project was abandoned, and the Token market value was less than US$10 million, a drop of 99%

This is not fiction, but the real trajectory of countless Web3 projects. In less than half a year, the project “returned to zero.”

3.2 The Trap of the Attention Economy

The Web3 industry has a unique phenomenon: attention is more important than product.

The success of a project largely depends on its ability to attract the attention of the community. And what's the best way to grab attention?

Create FOMO (Fear Of Missing Out).

Airdrop expectations, listing rumors, KOL orders, price surges... these can attract a lot of attention in a short period of time.

But this attention is speculative, not practical.

Users care about "can I make money", not "whether this product is useful".

OpenClaw's growth model is completely different. Its 150,000 GitHub Stars come from:

• Developers really find it useful

• Word of mouth in the technology community

• Ability to solve practical problems

There is no airdrop, no listing, and no KOL orders.

3.3 The paradox of "decentralization"

One of the core concepts of Web3 is "decentralization", but ironically, the Web3 project itself is often highly centralized.

• The core team controls the code repository

• Foundation controls Token distribution

• A few whales control voting rights

Although OpenClaw is not completely decentralized (Steinberger is the core maintainer), its open source nature means that anyone can fork, modify, and deploy their own version.

More importantly, OpenClaw does not have Tokens, so there is no complexity of economic incentives.

In Web3 projects, token holders, developers, and users are often different groups, and their interests often conflict.

In OpenClaw, these three people are the same group of people: the people who use it are the people who develop it, and the people who maintain it.

3.4 Fundamental differences in community culture

Let's compare the community culture of OpenClaw and a typical Web3 project:

OpenClaw Discord:

• Discuss how to deploy to Raspberry Pi

• Share custom Skills

• Report bugs and make feature suggestions

• Someone occasionally posts memes

Typical Web3 project Discord:

• Discuss currency prices and airdrops

• Ask "When will I go to the top law firm"

• Scolding the team for not pulling the strings

• Sending single messages and fraudulent links

This difference is not accidental, but the result of product positioning and community governance.

When the core of the project is Token, the community will naturally discuss around price.

When the core of the project is technology, the community will naturally discuss around the technology.

Conclusion: In the second half of Web3, it’s time to learn “ClawFather”

The story of Peter Steinberg reflects the divide between the two technical spirits:

One side is the spirit of a builder: I want to change the world, so I have to make useful things and solve real problems.

The other side is the spirit of speculators: I don't care what value the project provides, I only care about whether I can issue coins, whether I can exit, and whether I can make quick money.

OpenClaw was not born in Web3, not because Web3's technology is not good, but because Web3 focuses too much on how to get listed, how to CX, how to pull the market, and how to cut leeks. Real builders will only choose to stay away, just like what Steinberg wrote in server rules:

"No chatting about finance or crypto."

This is perhaps the sharpest criticism of Web3 and the most sincere advice.