Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
Singapore’s GIC, one of the world’s largest institutional investors, has advised “doubling down” on Big Tech rather than overvalued start-ups in the booming generative artificial intelligence sector.
Lim Chow Kiat, chief executive of the $700bn sovereign wealth fund, said tech giants, including Microsoft, Alphabet and Meta, had begun capturing “a lot of value” in the transformative technology with their existing customers.
“Customers are paying for [their services] already. So, actually, that is quite a good place for investors to double down on,” he said in an interview with the Financial Times. Valuations of start-ups with AI business models were “too high”, he added. “We need to be careful not to get sucked into the hype. But it is more important not to lose interest.”
Generative AI companies, whose technology can create imagery to order and humanlike text in seconds, have raised billions of dollars this year. Major tech companies have predicted that AI will drive their revenues, with Microsoft taking an early lead in incorporating it into its products, helped by its large stake in ChatGPT creator OpenAI.
GIC, which manages Singapore’s foreign exchange reserves, does not detail its investments or break down its portfolio by sector, but it does so by region. Its investments in the US rose to 38 per cent of its portfolio for the 12 months to March this year, against 37 per cent a year earlier, compared to a dip in Asia, excluding Japan, from 25 per cent to 23 per cent.
But it was one of the earliest institutional and sovereign investors to embrace tech, in an attempt to keep the tiny financial hub abreast of the latest innovations.
Of GIC’s $52bn invested in listed companies, $17.4bn is in the tech hardware and semiconductor sector, according to Bloomberg data. Its biggest current holdings in listed tech companies are Taiwan Semiconductor Manufacturing Company ($11.1bn), DoorDash ($2.1bn), Sony ($1.56bn) and MediaTek ($1.25bn).
Companies in its AI portfolio include UK analytics group Quantexa; Hong Kong data company MioTech; US AI software and hardware provider SambaNova Systems; and US-Israeli fintech Pagaya Technologies.
Both GIC, as well as Singapore’s other state investor Temasek, have described AI’s potential as transformative for business.
But the frenzy since the launch of the ChatGPT chatbot a year ago has prompted warnings from experts that most start-ups in the sector are overvalued and will fail to make money. AI has similarly stretched the market caps of big tech companies, with trillions added to their collective valuations since last year.
Venture capital firms and larger investors, including sovereign wealth funds, have watched from the sidelines as Big Tech has forged deeper relationships with the top generative AI start-ups. This year, Microsoft has invested billions of dollars more into OpenAI and backed rival Inflection. Alphabet and Amazon have committed up to $6bn to Anthropic.
The companies have been preferred partners because only they can provide the combination of computing power and capital required to build foundation models. In turn, they have received preferential access to AI tools, and are expected to recoup much of their investment from AI start-ups’ spending on their cloud services.
Infrastructure is another important area for investors, including GIC. “We believe there will be an infrastructure layer for AI the same way there was for the transition to the cloud,” said its chief.
Amazon, Microsoft and Google parent Alphabet have boosted investment in cloud computing infrastructure over the past few years. Bank of America analysts predict that the collective cloud-related capex of the three will rise at an accelerated 22 per cent next year to $116bn.
Singapore has taken a more collaborative approach with AI to date, rather than seeking to regulate it. In an annual letter in July on GIC’s performance, Lim warned that there were legitimate concerns that generative AI was being developed too quickly. He cited deepfakes, fake news and cyber attacks.
“Different investors have different ways of cutting the [AI] value chain, but . . . whatever way we cut [it], we have to remember that it is a long game,” he said.
Additional reporting by George Hammond in San Francisco and Mary McDougall in London
Read the full article here