Telegram, OpenAI, and the Market for Trust: Why Polished Numbers Can No Longer Be Taken at Face Value
Why can polished numbers from Telegram, OpenAI, and other technology companies no longer be accepted without verification?
Test headline technology metrics through counting methodology, ownership structure, audience quality, and the actual rules governing data.
What to watch for
Key takeaways
The practical meaning of “ToTheMoon — the best IT podcast is back on the air!” is that the audience grows, the product "changes the market", the next round will be even bigger — but what matters to investors and users is not the polished story, it is what hides behind the wording.
The working conclusion from “Telegram is unprofitable — where does the money for the product come from?” is that the conversation opens with an uncomfortable figure — an annual loss estimated in the hundreds of millions of dollars — and immediately returns to the old question: what is supposed to sustain the service, and who ultimately carries the economic risk.
The boundary of the “How big companies doctor their reports. The case of Elon Musk and X” case is defined by this point: private companies have almost no governance — FTX being the biggest reminder — but even public companies write reports so vague, with numbers so stretched, that everything is technically disclosed while the real picture stays blurred.
For the “Why did Pavel Durov get active in the public space, although it was incognito before” scene, the decisive point is this: even public-company reports can be written so that everything has technically been disclosed while the real picture remains blurred. After FTX, the market already knows what faith in status and founder charisma can lead to. Pavel Durov's public activity therefore becomes more than a personal style; it is part of the company's financial and political strategy.
The decision in “The battle of Google, Apple, and Microsoft for OpenAI, makers of ChatGPT” depends on one criterion: OpenAI was long seen almost as an extension of Microsoft, but Apple and NVIDIA appear in the new round: for Microsoft that weakens exclusivity, for NVIDIA it means sharing in the growth of compute's biggest consumer, for Apple a chance to embed the strongest models without building everything from scratch.
The boundary of the ““This content is created with AI”, why are they adding this button everywhere? What's the point” case is defined by this point: AI can be an ordinary tool in a person's hands or generate an entire image, voice, or video — the same label for both cases protects the platform legally more than it explains to the viewer what actually happened.
The “The legislative side of the question, what is Law SB 1047?” scene leads to a working conclusion: what is needed are understandable numbers, a clear ownership structure, rules for handling data, and an honest explanation of where the tool ends and the automated decision begins.
What this episode is about
Telegram's losses, ambiguous reports from large companies, OpenAI's new funding round, and mandatory labels for AI content look like unrelated stories. In fact, they are one conversation about trust: who controls the numbers, who owns the technology, and what exactly a company must explain to users and investors.
Technology companies like to talk about scale: the audience is growing, the product is changing the market, and the next funding round will be even larger. But users and investors need more than a polished story. They need to know what sits behind the language.
The discussion of Telegram's finances begins with an uncomfortable figure—an annual loss estimated in the hundreds of millions of dollars—and immediately returns to an old question: what is supposed to sustain the service, and who ultimately carries the economic risk?
The problem is not limited to privately held Telegram. Even public-company reports can be written so that everything has technically been disclosed while the real picture remains blurred.
After FTX, the market already knows what faith in status and founder charisma can lead to. Pavel Durov's public activity therefore becomes more than a personal style; it is part of the company's financial and political strategy.
At the same time, the balance around OpenAI is changing. For a long time, the company was perceived almost as an extension of Microsoft, but Apple and NVIDIA appear in the new funding round. For Microsoft, that means weaker exclusivity.
For NVIDIA, it is an opportunity to participate not only in chip sales but also in the growth of the largest consumer of compute. For Apple, it is a chance to bring the strongest models into its own ecosystem without building everything from scratch.
Against this backdrop, a label saying ‘created with AI’ looks like an attempt to restore at least some transparency at the content layer. But the label itself does not resolve authorship or responsibility.
AI can be an ordinary tool in a person's hands, or it can generate an entire image, voice, or video. Applying the same label to both cases protects the platform legally more than it explains to the viewer what actually happened.
All of these stories reduce to one rule: the more a company asks us to trust it with money, data, or attention, the less it can be allowed to live on brand alone. We need understandable numbers, a clear ownership structure, rules for handling data, and an honest explanation of where a tool ends and an automated decision begins.
All of these stories reduce to one rule: the more a company asks us to trust it with money, data, or attention, the less it can be allowed to live on brand alone. As a result, we need understandable numbers, a clear ownership structure, rules for handling data, and an honest explanation of where a tool ends and an automated decision begins.
Episode transcript
The episode is in Russian; below is an English reading guide to the transcript (the full EN transcript is a machine translation). Voice matching applied to 79 segments: 28 identified, 3 mixed, 20 probable, and 28 unresolved.
Read transcript on a separate page
Loading…