Is the hype for generative AI justified?

If you searched for the word “hype” in a dictionary published over the last few months, generative AI would probably be included in the entry.

Hype isn’t always a bad thing, and it can help develop real tech that addresses the great challenges that humanity faces. However, the negative side of the concept is still worth exploring.

For example, excessive publicity can create unrealistic expectations that end in disappointment, damaging reputations and stymying development. Gartner’s Hype Cycle illustrates this phenomenon, and the 1990s dot-com bubble and its thousands of bankruptcies is a good case study.

A visualization of the Gartner hype cycle / Chart by Jeremy Kemp

However, regardless of the hype and the billions of dollars of investments lost in the “Trough of Disillusionment,” the dot-com bubble changed our lives forever.

Despite the buzz around generative AI, there are hurdles for the tech in scalability: the cost of training neural networks, a lack of regulation, and ethical issues.

But in contrast to other hyped-up tech like the metaverse and cryptocurrency, the real applications of generative AI are apparent. For example, US-based edtech company Chegg lost half of its market cap after surrendering traffic to ChatGPT, showing that generative AI can already disrupt industries.

However, this alone doesn’t necessarily mean that all the noise is justified, as some expect the tech to bring sweeping societal change. So how do we differentiate between pure hype and potential?

Hype and high potential

As I see it, there are a few key measures we can use:

1. Impact: Will the tech have a significant impact on lives or businesses? What are the potential benefits?
2. Background: Is the tech based on underlying principles like sound science or engineering?
3. Viability: Is there a clear path to commercialization? Are there other major challenges, such as security and regulation, to overcome soon?
4. Drivers: What is driving the tech? Is it scientists and entrepreneurs or mostly influencers, journalists, and futurologists?

When we started focusing on cloud computing in 2010, the tech was a hot topic because it promised to significantly decrease IT costs, increase flexibility, and shorten the go-to-market cycle for new solutions. However, there was much skepticism overall and within our company, Runa Capital, due to concerns about data security and privacy.

At the time, cloud computing entered our investment thesis for the following reasons:

• The impact would be huge. Cloud computing would provide infrastructure for most upcoming tech businesses.
• The underlying tech such as high-speed internet and virtualization platforms were already in place.
• The economic effect of cloud computing was already proven and deployed in the early 2010s by big corporations such as Coca-Cola, General Electric, and Bank of America.
• Adoption was being driven by large companies like Amazon, Microsoft, and Google.

Generative AI, at this somewhat early stage, also seems to check all four boxes. The impact is already being felt, the tech is built with a sound background, and it’s being commercialized as well.

Avoiding the hype