Laurence Fletcher and Costas Mourselas’s report on Elliott Management and the advice the hedge fund gave its investors about Nvidia’s share price bubble is a refreshing read, echoing as it does the sentiment of many industry players, myself included (Report, August 3).

We are living in a real-world version of the “emperor has no clothes”. Visiting the US west coast recently, I found the noise levels around artificial intelligence overwhelming, and the promises around its applications reaching saturation point and full of false claims. We got to this point because of Fomo (fear of missing out). While advances in AI have been remarkable, we’re not even close to reaching the AI dream. Investors, companies and entrepreneurs have jumped on the bandwagon, scrambling to position themselves before any proper assessment of the claims put forward for this new technology can be tested. This is much to the delight of chip manufacturers and tech CEOs — hence Nvidia’s stock soaring and inflated valuations surrounding these “groundbreaking applications”. We are starting to see the tide turn. Elliott’s letter may seem dramatic but it brings up a lot of valid points. AI has failed to improve productivity. Research by America’s Census Bureau found only 5 per cent of businesses currently use AI. In June Goldman Sachs issued a damning report titled: “Gen AI: too much spend, too little benefit?”. Elliott is right to warn on chip stocks, with a further 5 per cent fall in Nvidia since the article and continued corrections across the majority of the Magnificent Seven tech stocks. Ultimately, no amount of manufactured chips or modelling will solve AI’s plateauing capabilities. By the same token, as more computing power becomes available, many are doubling down on large language models (LLM) — those AI systems capable of understanding and generating human language by processing vast amounts of text data. Some of us in the AI community have known for some time that there are gaps in intelligence that will stall AI’s leap from tactical workhorse to high-level decision maker in business. I do remain firm in the belief that AI will be the game changing technology and deliver value that would lead to mainstream adoption. But not in the hype we see today. We must take stock of the bigger picture, step away from the constant quest to build bigger models and ask if alternative architectures are another way of reaching AI nirvana. We are still only at the start of the journey towards super intelligence.

Imad Riachi
Founder and Chief Executive, Honu, London SW6, UK