Tag Archives: YouTube

In recent years, it has become routine for tech companies to overhype up their latest artificial intelligence technology. Ever since the rise of ChatGPT, the market for this technology has expanded faster than anything I’ve ever seen in my lifetime. It’s not unreasonable to surmise that AI will be a multi-trillion-dollar industry in the coming decades.

But I’m also old enough to remember the dot-com bubble. I haven’t forgotten what happens when hype exceeds capabilities. And, as remarkable as these emerging AI systems are, their capabilities are still limited. Whether it’s large language models or image generators, these AI systems are a long way from becoming Skynet. They’re certainly part of the ongoing development to create Artificial General Intelligence, but that leap is not imminent. That would be like jumping from rotary phones to iPhones.

However, AI is different from the dot-com bubble or any other overhyped product for that matter. I’ve said before that AI has the potential to change the world in ways we literally cannot imagine. That potential is far from being realized, but there’s a good chance people my age will live long enough to see it play out.

Well, recently a bit more of that potential was realized. Google, a company that is very active in the development of AI, released a new generative AI tool called Veo3. Like other generative AI tools, this particular tool specializes in creating full videos from text prompts. It’s not the first tool of its kind. However, based on some of the videos that have come out from Veo3, it’s definitely a step forward in terms of quality. If you need proof, check these videos out.

Remember, the content in this video is entirely AI. It never happened.

I’m not gonna lie. I did not expect these videos to look this realistic. With other AI video generators, you can usually tell it’s not real. Between the glitchy movements and the uncanny valley effect, not many people would mistake those videos as real.

But Veo3 has taken it to another level. If I didn’t know ahead of time that this content was fake, I might have just shrugged and accepted it. And that’s a major threshold that we cannot uncross with this technology. Because when it suddenly becomes possible to create videos that look as real as anything taken in real life, how can we trust any video?

Make no mistake. This won’t just be used to flood the internet with advertisements and mindless entertainment. Someone is going to use this technology maliciously. There’s a good chance someone is already doing so as we speak.

At this very moment, you can use Veo3. It is expensive, as many advanced AI tools tend to be. But if you’re determined to make a certain video for a certain purpose, be it malicious or altruistic, is $250 really that high a price? I know people who have spent more on skins in Fortnite.

And Veo3 is not the apex of generative AI. It’s going to keep improving. The incentives are too strong. Google might have the best video generator now, but other multi-billion-dollar companies who want a piece of that AI market will fight to take that title.

Maybe OpenAI will upgrade Sora.

Maybe Microsoft will upgrade its Copilot AI.

Maybe a company like Apple will make a big leap, just like they did with smartphones years ago.

But no matter which company makes the next best AI video generator, the technology will improve. It’s going to become increasingly difficult to tell when something is real or fake. There are likely going to be laws and policies that attempt to regulate this kind of AI. But governments have been historically slow at keeping up with technology. And the advancement of AI is likely to accelerate.

I certainly don’t claim to know what kind of impact Veo3 will have in the short-term. I already make YouTube videos on my own. Will my kind of content become obsolete or unable to compete? Will video sites everywhere become overwhelmed or bloated because of content generated by Veo3?

Then, there’s the long-term impact. What happens to major studios once this technology allows them to make movies and shows faster, cheaper, and without paying actual people? That’s an issue I once thought was decades away. Now, it might be relevant within a few years.

Only time will tell what sort of changes this technology will incur. But if what Veo3 has demonstrated is any indication, these changes are coming a lot sooner than expected.