This is a clickbait title. Here’s what he actually said:

Clickbait_Tweet

Yes, it’s intended to be that way so people read it and react to it. I know it, so does everyone that writes articles like these.

And people fall for it. Not just your average joe, but extremely well-read people. So has been traditional media for decades and surprisingly, so are platforms like LinkedIn.

If you’ve been on social media in the past week, you’ve seen people react to this tweet by Sam Altman:

Electrcity_costs

what is extremely funny is that many self proclaimed “AI Experts”, Founders and people with Executive level positions in a lot of major places put out stuff like this (literally didn’t even sort by latest, just searched OpenAI here)

linkedinpostslol

What. The. Actual. Hell.

People do not understand what’s a joke, and it’s extremely concerning because of the fact that we’re not in a traditional exclusive media format right now where narratives can be changed based on the person or the author presenting it. This is social media.

A guy on X actually asked ChatGPT o3 to estimate said costs, and as expected, it’s less than 150k (which is nothing on the scale OpenAI operates).

Image of calculations

credit - @luismbat on X

We’re nerds, we make jokes, we mess around. We put a 5’7” guy who laid across the Harvard bridge as a unit of measurement on Google Earth (It’s called smoot, seriously check it out). And it surprises me how “professional” platforms haven’t caught up to that fact.

The Ever-Existing Clickbait Economy

This phenomenon isn’t unique to AI or tech news. The pattern of sensationalizing statements, removing context, and manufacturing outrage has been a media strategy for generations. You see it all the time on YouTube, But what’s different now is the speed and scale at which misinformation spreads.

You know what this AI scene somewhat draws a parallel to though? The dot-com bubble. During the dot-com bubble of the late 1990s, traditional media played a similar role in both hyping and then demonizing internet companies. Headlines would swing wildly from “The New Economy Will Change Everything Forever” to “Internet Stocks Worthless, Experts Say” within months.

The dot-com era saw media outlets breathlessly reporting on companies with no revenue and absurd valuations, helping fuel what would become one of the most extreme speculative bubbles in market history. Publications that once celebrated these companies as revolutionary later mocked them as obvious failures after the crash.

The Dot-Com Media Circus and Today’s AI Hype

The media coverage of the AI boom now is kinda similar to the whole dot-com thing back in the day. It’s like history is repeating itself! The news always goes to extremes, showing only the super exciting or scary sides of things. It’s like they take what folks say and twist it to make it seem way more dramatic than it is. Also, loads of people are claiming to be “experts” when they don’t really know that much about AI. It’s like everyone’s all hyped up about the wrong stuff, focusing on the wrong metrics and numbers. They’re just different from the old ones, like “eyeballs” and “clicks” have become “parameters” and “tokens”.

The Media’s Role in Bubble Creation

The media has this interesting relationship with technology bubbles - they don’t just report on them, they actually help make them! Back in the dot-com days, business newspapers like The Wall Street Journal and magazines like Forbes extensively covered “the tech boom”. They made it seem like investing in startups was the best idea ever, fueling the bubble.

And guess what? We’re seeing the same with AI now. The media is going wild over AI, with headlines everywhere shouting about the next big thing. Everyone’s getting FOMO (fear of missing out), which makes them invest more and more money in AI startups. Venture funding in this area has gone through the roof! It’s like a self-fulfilling prophecy: the media hypes AI, which makes everyone rush to invest, which then confirms the hype. It’s a viscous cycle.

The media’s role in both eras follows a predictable pattern which I like to call CLOWN

  • Captivation (Discovery phase): Breathless coverage of a new technology’s potential
  • Laudation (Evangelism phase): Uncritical amplification of the most optimistic voices
  • Obsession (Gold rush phase): Detailed tracking of funding rounds and valuations, creating FOMO
  • Waning (Disillusionment phase): Sudden pivot to skepticism when the bubble shows signs of weakness
  • Negation (Blame phase): Articles explaining why the bubble was obviously unsustainable all along

We’re currently somewhere between phases 2 and 3 with AI, with some early signs of phase 4 emerging as certain AI startups fail to deliver on their promises.

The Speed of Misinformation

What’s different now is the velocity. During the dot-com era, a misinterpreted statement might take days to spread through traditional media channels. Today, a single post can generate thousands of reactions within hours, with each layer of sharing further removing context and nuance.

The AI boom, like the internet boom before it, represents a genuine technological transformation. As one analysis put it: “The late 1990s marked a surge in internet adoption, with global internet users rising from 16 million in 1995 to over 300 million by 2000, fundamentally altering communication and information access.” Today’s AI advancements are similarly revolutionary, but the noise-to-signal ratio in media coverage makes it difficult to separate substance from hype.

The Bubble Economics of AI

The way companies and startups are valued in tech bubbles is pretty interesting. In the dot-com era, companies were valued based on how many people visited their websites, which doesn’t really make sense when you think about it. Nowadays, AI startups are valued based on the size of their data or how well their models perform on benchmarks. But do these things really tell us if the company is going to be successful?

The stock market back then saw huge rises and falls that seem similar to what’s happening now with AI. The costs of developing AI models have dropped massively, and AI startups are losing loads of money just to stay afloat and capture a part of the market. But the news we hear about AI is all super positive, focusing on how amazing it all is rather than whether it’s actually a good investment. History seems to be repeating itself!

Learning from History

The dot-com bubble ultimately burst on March 10, 2000, wiping out trillions in market value. Many companies that had achieved market capitalizations in the hundreds of millions became worthless within months. The aftermath was devastating, with massive layoffs and investor capital evaporating overnight.

Yet from those ashes rose the tech giants that now dominate our digital landscape. Amazon, eBay, and others survived because they had sustainable business models beneath the hype.

It’s the same old cliche: Don’t believe the hype. The loudest narratives out there aren’t always the most reliable when it comes to understanding tech. We should probably be looking closer at the companies and technologies that aren’t trying to grab all the attention. They might just be the ones that last.

The Solution?

There isn’t one simple fix. You, as a reader, are responsible for knowing and verifying what you read. Find the source, check it, see if you interpret it the same way.

The internet can be a very hard place to be if you don’t understand humor or sarcasm. That’s the truth.

As we navigate the current AI boom, remember that headlines are designed to provoke emotion, not convey nuance. Just as the dot-com era eventually separated sustainable innovations from speculative excess, the AI landscape will eventually be defined not by tweets and hot takes, but by technologies that deliver genuine value.

In the meantime, maybe we could all benefit from saying “please” and “thank you” to our AI models especially because it doesn’t actually cost tens of millions of dollars…