The latest destination on the AI king show hosted by Sam Altman, the CEO of OpenAI, was the United Arab Emirates. During his visit, Altman proposed the concept of establishing an international coalition comprising private entities and governments to foster and sustain the artificial intelligence sector.
Altman engaged in discussions with Emirati officials and traders to explore collaborative strategies aimed at mitigating the substantial costs associated with AI systems. The agenda encompassed aspects such as energy supply, transistor procurement, and server capacity. Furthermore, Altman met with Martina Strong, the US ambassador to the UAE, as part of his visit.
In an effort to establish a network of chip manufacturing facilities focused on catering to the needs of machine learning, Altman reportedly solicited funding amounting to billions of dollars from industry behemoths like G42 based in Abu Dhabi, Softbank from Japan, Microsoft, and others over the past few decades.
These deliberations extended beyond merely augmenting the proliferation of AI startups that drive initiatives like OpenAI. They also delved into accelerating server expansion and leveraging alternative energy sources, including nuclear power—a domain where Altman has demonstrated significant advocacy, previously endorsing startups like Oklo specializing in small modular reactors and Helion Energy focusing on fusion technology.
OpenAI, in a statement to Bloomberg, hinted at ongoing dialogues concerning the amplification of global infrastructure and resource repositories for chips, power, and data centers, with more comprehensive details slated for disclosure in the future.
The UAE, known for its substantial financial resources and keen interest in AI investments, has been actively pursuing ventures in this domain. However, concerns have been raised regarding the nation’s track record on human rights. Notably, the G42 investment group within the UAE has allocated substantial sums towards AI infrastructure projects, including a notable initiative involving Cerebra’s wafer-scale accelerators, entailing a $900 million investment.
Recent scrutiny has surrounded G42’s affiliations following allegations of supplying advanced AI systems and genomic data to China, detailing information on millions of individuals. To assuage concerns, G42 purportedly severed ties with Taiwanese vendors.
Apart from engagements in the UAE, Altman has also conducted similar discussions with European nations. Speculation suggests that the implementation of his proposals could entail costs ranging from billions to trillions of dollars, with reports indicating Altman’s purported quest for over $7 trillion for semiconductor fabs—an amount significantly surpassing the total revenue of the semiconductor industry in 2023.
While Altman has refuted claims of seeking trillions in funding, he emphasized the substantial financial backing required on a global scale to advance AI technologies. Acknowledging the underestimated expenses associated with AI advancements, Altman highlighted the necessity for extensive AI computing resources.
Despite efforts to ramp up production, the demand for power-intensive GPUs and accelerators essential for training and deploying large AI models continues to outstrip supply. Analysts anticipate a persistent supply-demand gap in the foreseeable future, even as Nvidia, a prominent AI systems manufacturer, aims to substantially increase the production of its H100 and H200 accelerators this year.
