Who wants to be a billionaire?

What happens when everyone can start their own company? Anthropic CEO believes almost all code will be AI-written by the end of 2025, and OpenAI is developing a new device that will retire the smartphone.

Welcome to Kludder of the week!

Can a company valued at one billion dollars be run by just one person? Anthropic CEO Dario Amodei certainly thinks so! During Anthropic Developers Day last week, which gathered over 500 guests in San Francisco, Amodei spoke enthusiastically about how artificial intelligence will change the way we build companies. He's convinced that as early as 2026, we'll see the world's first billion-dollar business without employees.

Dario Amodei isn't alone in his lofty expectations. In February 2024, OpenAI CEO Sam Altman said something similar, though without a specific year. Altman believes it's likely that a creator, will be behind the billion-dollar milestone.

AI can replace "everything"

I get what Altman means. It's no longer utopian to think that a weekly newsletter like Kludder could be transformed into a podcast, broken up into social media content, or rewritten as op-eds and debate pieces. With just a few keystrokes, I could create presentations based on the newsletter, or use AI to generate pitches for advertisers. With Google's new video AI, VEO 3, I could create YouTube clips with my face, my voice, and Kludder's logo.

Suddenly, a weekly newsletter seems terribly analog...

But the point is that both Altman and Amodei expect AI to change how entrepreneurs scale businesses. Today, you depend on teams with different expertise - product designers, sales managers, marketers, developers, and HR functions. But soon, AI could replace all these. Content production, data analysis, and customer service through AI bots could provide lightning speed toward profitability, without hiring a single human being.

In our little group chat with my tech CEO friends, we have a betting pool on the first year there will be a billion-dollar company with just one person. - Sam Altman

The tools exist

There are companies that promise to give you the tools to build human-free businesses. Relevance AI, with offices in Sydney and San Francisco, is building what they call an "operating system" for AI agents. The idea is that you can deploy agents that specialize and adapt to different roles. The company recently raised 37 million dollars and plans to compete with Microsoft, among others, which is also developing AI agents. The market around AI agents is expected to become highly lucrative: A study from Boston Consulting Group estimates that the AI agent market will grow by approximately 45 percent annually over the next five years.

Who are they talking to?

Amodei's statements have received a lot of attention, and every time Altman opens his mouth, the media pounces on it. But how can the AI executives predict with such confidence what will happen in 2026?

In addition to Anthropic and OpenAI, there are other language models on the market. xAI is Elon Musk's venture - with varying success. Meta has Llama, Microsoft has Copilot, Google has Gemini, and then there's the Chinese DeepSeek. It's getting crowded in the AI race. When it's also enormously costly to be at the forefront of language model development, there's an eternal battle for investors' attention. Therefore, I wonder if statements like those from Amodei and Altman are equally investor pitches as actual claims:

Within three to six months, 90 percent of code is written by AI - Dario Amodei, three months ago.

All bets are off

So far, OpenAI has dominated the "AI market" with aggressive marketing of new features on X, where every minor improvement is launched with fanfare. Anthropic has taken the opposite approach, working in the shadows and being restrictive with launches. But with their first Anthropic Developers Conference, it appears that the company, now valued at more than 61 billion dollars, is done with walking quietly in the hallways.

Regardless of whether Altman and Amodei are courting investors through clever media attention, or actually believe their own statements, we're facing a dramatic shift. Last week, I wrote about OpenAI's acquisition of hardware company io, an acquisition worth over 6.5 billion dollars for a company housing only 55 employees.

Mike Krieger, co-founder of Instagram and now CPO at Anthropic, commented on Amodei's predictions:

It's not completely crazy. I built up a billion-dollar company with just 13 people.

Instagram's challenge eventually became moderating offensive content, which required more hires. Today, that challenge could have been solved by AI.

The efficiency paradox

If AI is to replace all these functions and roles, the job market will be completely turned upside down. Millions will become unemployed. Or they'll become entrepreneurs, but if anyone who wants to can start a billion-dollar company, who will become their customers?

And perhaps this is the point: We're not just facing a technological revolution, but a fundamental change in what a company and a business really is. When production costs approach zero, efficiency shoots through the roof, and distribution becomes free, value creation must be measured in a completely different way.

It remains to be seen whether Altman and Amodei's visions become reality - or if they remain lofty investor sweet-talk.

In the meantime, we can all just vibe-code our way to billionaire status.

Enjoying Kludder? You're not alone! Many people are reading the newsletter every week - but we've got room for more of you. Feel free to share us with a friend or colleague.

Stuck in 2024

AI Overview is Google's AI-generated summary that appears when you google something. Both Wired and The Verge report that Overview thinks we're still in 2024. Among the sources it cites are a Reddit thread and a Wikipedia page. When I did my own tests, Overview appears to be updated.

The New York Times agrees with Amazon

The New York Times sued OpenAI and Microsoft for copyright infringement in 2023. The newspaper believed the companies had trained their AI models on millions of articles without NYtimes' consent.

Now Amazon has secured rights to use the newspaper's articles on their products, including the smart speaker Alexa. I think we'll see more of these types of agreements, where media houses and newspapers let companies train artificial intelligence on their content. For a handsome fee, of course.

Without a trace

An interesting article from The Atlantic highlights the company Privacy.com. By subscribing to Privacy.com, you get access to virtual bank cards with fake names to prevent you from leaving digital traces. In the article, you can read about Alec Harris, who has taken his own privacy to extreme lengths.

Less screen time?

Last Thursday, the Wall Street Journal held its Future of Everything conference. In conversation with journalist Keach Hagey, OpenAI's Chief Operating Officer Brad Lightcap says the company wants to develop a device that will make us "less dependent on the screen".