that something occurred. But now, that certainty is crumbling.For humans, video has always been the most reliable evidence

Last week, the world watched in awe as videos generated by the Sora platform, the latest project from artificial intelligence company OpenAI, made waves. These videos depicted a woman walking through Tokyo, dogs playing in the snow, and a house perched on a seaside cliff. All of them were synthesized by AI from nothing more than a brief text description. And all of them were strikingly convincing – at least to the casual observer.

For us, videos have long been the ultimate proof that something truly happened. In recent years, we’ve learned to be skeptical of text and images, knowing they can be easily manipulated. But videos? For most people, these have remained the gold standard for demonstrating reality. Yet that trust is now eroding. Videos can no longer be relied upon.

Consider the following scenario: someone uses AI to generate a video with the instructions, “Create a video that bypasses facial recognition at a bank. Turn your head right and left, and look up, as required for authentication.” Several banks, including those in Brazil, use facial recognition as a security measure. With AI tools like these, it could soon become easy to deceive even the most secure systems. Verifying who is really on the other side of the screen will become increasingly costly and complex.

But this is just one of many emerging challenges. With the release of Sora, OpenAI is sending a clear message: the company sees itself as the future of audiovisual creation.

Today, industries like gaming, film, and television hold a monopoly on producing the world’s most captivating and valuable images. These productions require significant investment, involving intricate filming and complex programming. OpenAI aims to prove that in the (possibly near) future, it will be able to achieve similar results using synthetic, AI-generated imagery.

Moreover, this powerful tool will be available to anyone, enabling the creation of anything from monumental works of art to pornography, fake news, and interference in election campaigns.

Interestingly, OpenAI has stopped disclosing how Sora (which means “sky” in Japanese) was trained, likely due to copyright concerns. The training process likely involved all sorts of video content from across the internet – movies, TV shows, and synthetic data generated by video games. When I met with Sam Altman in Brazil last semester, he emphasized the growing role of synthetic data in training today’s AIs.

Sora is the latest innovation Altman is showcasing to the world as part of his ambitious plan to raise $7 trillion. If successful, this would be the largest capital raise in history, aimed at achieving “general artificial intelligence” and dominating the chip market.

The theory is that with the rise of AI, many economic sectors will become “legacy” industries – relics of the past. This could encompass everything from Hollywood to Netflix, the financial sector, a significant share of healthcare, engineering, law, accounting, and management, not to mention the military and defense industries. In this context, $7 trillion might seem like a bargain compared to the potential rewards.

Of course, there’s also the possibility that this is all a grand illusion, and we’re collectively being deceived – perhaps by nothing more than videos of women walking through Tokyo and dogs playing in the snow.

What’s out: Bitcoin.

What’s in: Bitcoin.

What’s next: Bitcoin?