I have test-driven countless models, but I always find myself coming back to Claude. Part of my job involves tracking inventory for equipment delivered to our company and compiling it into reports. I recently built an automated pipeline using Claude that organizes the raw data, drafts the document, and even emails it out automatically. Thanks to that, I get to clock out early these days.
We live in a world where a few clicks can generate code, draft professional documents, and translate languages in real time. Yet ironically, the company feeling the most anxiety in this booming era is the very pioneer that started it all. OpenAI.
The Financial Reality Behind the Hype
I have written about OpenAI's troubling situation before, and the storm has only grown fiercer since.
Following a staggering $7 billion deficit last year, rumors of bankruptcy started circulating, leading many to conclude that OpenAI's golden age is already behind it. The bleeding is expected to double this year to roughly $14 billion, and most analysts now project that turning a profit before 2030 is highly unlikely.
The most critical threat to OpenAI is its underlying revenue model. The harsh truth is that OpenAI has never once posted a profit. Revenue is climbing, sure, but the costs of research, development, and server maintenance are climbing even faster. It has become a leaky bucket. On top of that, the fierce talent war among Big Tech means they are handing out roughly $5 billion just in stock compensation to keep employees from walking out the door.
The math simply does not work. Right now, OpenAI is spending $1.69 for every single dollar it earns.
To patch this financial hole, CEO Sam Altman recently introduced advertisements, a move he once called a last resort. But this path is proving difficult. Ads might offer a quick revenue boost, but users are pushing back, questioning whether the AI is giving them genuine, objective answers or just a message shaped by who is paying for visibility.
Losing the Crown to Focused Rivals
Adding fuel to the fire is the relentless rise of competitors.
 |
| Market share growth from Sept 2024 to March 2026 |
Just a year ago, ChatGPT commanded an overwhelming 86% of the market. In the span of a single year, that dominance has dropped to 56%. The web traffic numbers tell an even sharper story. While Claude and Gemini saw their traffic jump by 770% and 800% respectively, ChatGPT managed a comparatively sluggish 84% growth over the same period.
I get why this is happening. Like most people, my AI journey started with ChatGPT. But as the ecosystem expanded, I found myself instinctively reaching for Claude whenever I need to write professional documents or do any coding, simply because it performs better. When I am working inside the Google ecosystem, Drive or Gmail, I switch over to Gemini without thinking twice.
Among my circle, a clear pattern has emerged. Use Gemini for search, Claude for coding. ChatGPT is left without a clear, defining strength.
At the end of the day, AI companies need to make money, which means convincing people to actually pay for a subscription. For any business, hesitation at checkout is fatal. ChatGPT's pitch has essentially been "our AI can do everything." But for actual users, an AI that is decently good at everything without a sharp specialty is no longer compelling.
The numbers back this up. As of March 2026, out of ChatGPT's roughly 1 billion weekly active users, only about 50 million are paying subscribers. A conversion rate of just 5%.
The Power of Selection and Concentration
In stark contrast to OpenAI's struggle for profitability stands Anthropic, the company behind Claude.
You could argue Gemini has the advantage of Google's backing, but Anthropic started as a scrappy startup, just like OpenAI did. And yet they are projected to post their first ever profit this very quarter. So what is the secret?
 |
| How Anthropic’s strategy is outpacing the competition |
Unlike OpenAI, Anthropic chose selection and concentration. They realized early on that trying to be the everything AI was a losing battle. Instead of burning massive amounts of money chasing a versatile, all knowing model, they sharpened one weapon and sharpened it well: coding AI.
That focus paid off. Leveraging this strength, Anthropic secured 8 of the top 10 Fortune companies as enterprise clients, and they are now expanding quickly into the everyday consumer market too.
From my own experience, the difference is tangible. My work often requires building complex Excel spreadsheets, and Claude consistently nails the exact formatting I need with better accuracy than anything else I have tried. With ChatGPT, I used to notice that once a conversation ran long, the responses would start losing context and degrading in quality.
Claude holds onto context no matter how long the conversation runs, and it wraps up tasks smoothly and reliably every time.
That reliability genuinely gives me peace of mind, which makes the subscription fee feel more like a good deal than an expense.
The Race for the First AI IPO
Backed into a corner, OpenAI pulled out what was supposed to be its trump card: an Initial Public Offering. The plan was straightforward. Capture the title of the first AI IPO and use that prestige to pull in a massive wave of capital to ease the cash crunch. Even with shaky financials, being first in a historic market shift tends to guarantee a flood of investor interest, which is exactly why they rushed to file.
But in a brutal twist, Anthropic beat them to it.
On June 1st, Anthropic submitted its IPO paperwork, officially entering the market a full week ahead of OpenAI. And the valuations made the blow even worse. OpenAI was valued at $852 billion. Anthropic came in higher at $965 billion.
We are watching the original pioneer of the AI boom face a genuine existential threat from the very competitors it helped inspire. Can OpenAI claw its way back and make the world say ChatGPT is still the king? I would love to hear what you think in the comments.
댓글 쓰기