The flame of AI mania that has been blazing in the cryptocurrency market appears to be flickering out. Once the darling of venture capital funds, particularly during the troublesome crypto times of 2022, AI large language models (or LLMs) are now showing signs of being a bubble that’s on the brink of bursting.
Much of the hype started around OpenAI’s ChatGPT launch in late 2022. With the crypto industry struggling, venture capitalists found AI technology an attractive refuge for their money. ChatGPT had reportedly drawn a million users within five days of being live, raising hopes among tech giants like Google, Microsoft, and Facebook, who swiftly entered the scene to exploit the emerging market.
Jasper AI, among numerous others, capitalized on this optimistic climate. The GPT-based copywriting platform achieved a $1.5 billion valuation after raising $125 million. In total, more than $25 billion has been funnelled into AI startups within the first half of 2023.
However, an ominous question lurks beneath the impressive numbers; where is this money going? Normally, funds raised would be distributed to research and development and customer acquisition. However, in the AI sphere, the funds are heavily directed towards securing high-performing graphics processing units (GPUs). These units are in such demand that a severe shortage has mounted, leaving AI startups fighting fiercely to out-pay each other.
A concerning matter is the reliance of some AI startups on APIs to siphon off services from OpenAI. While easier and quicker, this approach often means giving up control or ownership of intellectual property, which can lead to severe unintended consequences. Jasper AI and others like it have learned this harsh lesson, having to lay off a significant part of their workforce as demand for AI platforms decreases.
The market saturation is also evident in the declining interest in ChatGPT and other major platforms since their introduction. In fact, a recent Stanford study disclosed that GPT-3.5 and GPT-4’s performance on various tasks have deteriorated ‘substantially’ over time.
Ultimately, while core products like GPT and DeepMind are vital catalysts in the AI race, there’s an undeniable sense of a bubble being inflated due to the intense, perhaps overblown hype around the technology. All it would take is a couple of negative industry reports or bankruptcies to prick this bubble.
Despite these alarming observations, it appears investors and tech enthusiasts alike should approach AI technology with a cautious optimism, keeping an eye out for the tell-tale signs of instability amidst the market euphoria.
Source: Cointelegraph
