Parallels between the AI Boom and the Dot-Com Era, Including Investment Trends

Key Insight

The current AI boom exhibits striking parallels with the dot-com era, characterized by overheated rhetoric and grand, often utopian, promises. Similar to the internet's initial promise to connect the world and simplify lives, AI is touted as a solution for climate change, a means to reduce global inequality through AI tutors and doctors, and a pathway to a future where virtual assistants and robots alleviate most work, potentially leading to widespread ennui. Both technological disruptions are seen as 'before-and-after' moments, with an unclear future, and have generated belief among optimists that these periods mark the most transformative in human history, echoing sentiments from 1995 to 2023.

The rapid emergence of AI, much like the internet, followed years of advancements, reaching a tipping point where it seemingly appeared out of nowhere. This sudden ubiquity led to AI-generated content, such as a stylish image of the pope or a fake Joe Rogan, becoming social media sensations, and an Indian artist's 'Slumdog billionaires' series depicting tech magnates in poverty. This fervor quickly extended to the investment landscape, reminiscent of the dot-com startup frenzy where firms made million-dollar offers on the spot without due diligence. The AI sector saw similar overexuberance; for instance, Mistral, founded by three research scientists with just a seven-page memo, raised 105 million euros, approximately 117 million dollars, with future funding rounds pushing its valuation to over 6 billion dollars, reflecting a 'gold-rush mentality' where investors 'throw money' at ideas.

Business plans were often ephemeral for AI startups, similar to the dot-com era, with companies like Inflection requiring immense capital but lacking concrete revenue generation strategies. A 'freemium' model became prevalent, exemplified by OpenAI, where subscribers paid 20 dollars a month for advanced models like GPT-4, while earlier, less powerful versions remained free. The significant performance difference, particularly in tasks involving reasoning and explanation, was starkly apparent upon using the upgraded models. This explosive growth was tracked by platforms like 'There's an AI for That,' which listed over 3000 startups in March and added an average of nearly 1000 new companies monthly through the end of 2023, equating to about 30 new entities daily.