Much has been made of the generative AI boom in recent months. It seems like every tech company, VC and founder is exploring something related to AI. It also seems like we’ve been here before. Less than 12 months ago, a similar frenzy was fueled by innovation in web3. Now all we hear about is FTX, Sam Bankman-Fried and a nightmare of financial unwinding to come. Yet it’s not too difficult to separate what is happening in AI now, from web3 then. There are three reasons why the hype over generative AI is warranted and sustainable. 

Real, tangible utility

1. The success of any consumer product lies in its ultimate utility to the consumer. The biggest and most important difference is that AI technology has already demonstrated tangible benefits in the form of efficiency gains. Think about the many things we use the internet for today, the potential of generative AI applications is that they can utilize all this data and make what we want to do much faster.
2. Already, generative AI can be used to help us write and translate faster, it can create new assets like voices, images and videos with less time, and it can search the Internet and find us the answer more succinctly. It’s solving a very real and daily problem of “I don’t have enough time”. It’s making our version of the internet a faster and more efficient one.
3. In contrast, web3 has much wider and ambitious goals of disrupting the current version of the internet (hence the moniker web 3.0). But before it could get there, a lot of what was being built with crypto and blockchain technologies appeared to go the way of founders building a product and searching for its utility (and consumers) later. The most promising use case was non-fungible tokens, which essentially made digital assets immutable. But they were also known as expensive jpegs, because there was no obvious value of owning an NFT other than the ownership itself. The other goal was to provide an income stream for artists, but for this to be achieved, consumers had to believe in the first goal – that ownership mattered. It’s not clear they did. The scarcity value obscured its lack of real, tangible utility. 
4. OpenAI’s ChatGPT has become the fastest product to reach 100M monthly active users ever – in just an incredible 2 months. The biggest resource for any human is time. And if they do not find that a product is useful enough to trade their time for, they will not use it. 

Consumer familiarity and ease of use

1. Artificial Intelligence has been in the mainstream consciousness for a while already, whether in the chatbots we interact with, or the algorithms that show us what video to watch next on TikTok and Netflix. Even in popular culture, sci-fi dystopian movies and TV shows, as far back as Bladerunner, I, Robot and, more recently, Westworld have brought to mainstream consciousness the opportunities and perils of AI, and Artificial General Intelligence (AGI). This broad population awareness of AI has made the adoption of generative AI technology much faster because the mainstream consumer is not required to first learn and accept what the technology is, they already do. 
2. This is important in the innovation adoption curve, where new ideas are not adopted by all individuals at the same time. Instead you’ve got the early adopters, early majority then the late majority and laggards. The goal of any successful consumer product is to get to the early majority and late majority as soon as possible to achieve scale. Often after a certain point, consumer products hit a bump, and require marketing and broad consumer education to continue to gain new users. I think consumers broadly accepting that AI is, and will increasingly be a part of our society (we will allow robots to do certain jobs for us if it can be done faster) has given generative AI an advantage in the consumer education that web3 did not have. Therefore I can see generative AI consumer products going from the early adopters to early majority quickly. 
3. Comparatively, the technologies and principles underlying web3 are harder to grasp. The idea of a crypto economy began with the creation of Bitcoin by the mysterious Satoshi Nakamoto more than a decade ago. That led to blockchain and the potential to develop trustless and decentralized systems. But few people and companies were really ready to adopt that technology. Crypto stayed in the “early adopters” stage for a long time and only a few years ago started to move into the early majority stage.
4. The challenge for crypto, and more fundamentally, blockchain technologies, was that it required consumers to understand many concepts that were new and complex. While with generative AI it is easy to understand how it can help create new content and improve our efficiency, web3 use cases required consumers to believe that disrupting the current internet structure and its large centralized intermediaries like Google and Facebook, as well as financial institutions, would be better for consumers as a whole. 
5. This is not something easy to understand, nor is there widespread consensus on it. Instead a lot of consumer attention centered on making money from crypto tokens, which distracted from more purposeful innovation that was at the heart of web3 technologies. Crypto has not evolved yet from being perceived as just an alternative financial asset to a broader technological innovation.

Network Effects

1. Finally, like many successful scaled consumer products, generative AI products benefit from network effects. The more users who use ChatGPT or Dall-E, the more data the models will have to train on, and they will get better and better. Whether it is text, voice synthesis or design and art, the intrinsic value of generative AI models will be that they will get smarter and smarter, the more that people use it. At the same time, this will make the resulting efficiency higher and therefore create more utility for the consumers. This creates an incentive for people to keep using the product over and over again.
2. In contrast, web3 consumer products targeted small, exclusive communities. What it traded off in scale, it hoped to make up through token monetization and distribution. The scarcity value of tokens is good for the artists who create them and the owners who want to invest and make money from them, but it does not necessarily lead to scale. This is not to say that there will not be any large scale web3 consumer products in the future, but the fundamental nature of decentralized systems implies that it is better to be on many decentralized, small networks and not just one central platform, therefore implicitly network effects, and scale will be harder to achieve and sustain. 

The key tenets of successful consumer products underlie generative AI products: tangible utility, consumer acceptance, and network effects. These were not as apparent in the last consumer hype cycle of web3, but are what makes this hype cycle of generative AI products credible and sustainable. If you weren’t paying attention before, it’s time to get on board. As a consumer and investor, I am excited to see what’s to come.