The past two years of artificial intelligence (AI) have been like science fiction, but now it needs to move to the next stage, namely agentic AI that can work for us while being completely unobtrusive. Meanwhile, AI threatens to undermine our autonomy, even providing the ability to clone our identities and personalities.
Those were some of the views put forward at this week’s Leap 2025 tech show in Riyadh, Saudi Arabia, at which the kingdom announced almost $15bn worth of planned investment in AI.
The announcement came in the context of Saudi Arabia’s multi-year plan Vision 2030, which seeks to diversify the country away from its historic heavy dependence on oil production revenues.
Saudi Arabia is placing big bets on IT, datacentre capability and AI in particular, with the latter a massive focus of the show and an agenda packed out with discussions around applications of AI and its next steps, particularly agentic AI.
Projects announced in AI included a $1.5bn agreement between AI infrastructure provider Groq and Aramco Digital – a subsidiary of long-established state oil company Aramco – to expand AI-powered inference infrastructure and cloud computing. Also, Saudi state-owned manufacturing conglomerate ALAT and Lenovo committed $2bn to establish an advanced manufacturing – including semiconductors – and technology centre that integrates AI and robotics.
Others included Google’s new AI-driven digital infrastructure and the launch of a computing cluster, and Qualcomm announcing availability of its ALLAM Arabic large language model (LLM) on Qualcomm AI Cloud.
The kingdom also revealed that $42.4bn had been invested in technology-related infrastructure since 2022.
These included Databricks investing $300m in integrated platform as a service (PaaS) for developers in AI tools; SambaNova committing $140m to build advanced AI infrastructure; Salesforce investing $500m to develop Hyperforce and enhance cloud capabilities for regional customers; and Tencent Cloud allocating $150m to establish the Middle East’s first AI-powered cloud region.
With a couple of years of generative AI behind us, most notably in the form of ChatGPT, a big theme at the event was the next stages in AI development and discussions of its longer-term effects on humanity.
Of the former, agentic AI is touted as one of the next big things in AI. Here, the challenge is to move beyond AI as something we seek out and use, but instead works as a barely noticed adjunct to human activities that operates for us.
That was a view expounded by Yaser Al-Onaizan, CEO of the National Center for AI in the Saudi Data and AI Authority (SDAIA).
“The promise of AI is that it will be in everything that we do and we touch every day,” said Al-Onaizan. “It needs to be invisible. It cannot be in your face – it should be listening to you, understanding you and doing things based on your opinion.”
Al-Onaizan said the new generation of models is about doing something on your behalf. For example, instead of just giving you information about flights, it can go on and reserve flights, for example, or make a hotel reservation for you.
But some speakers warned of the threats posed by AI. Among these were Lambert Hogenhout, chief of data, analytics and emerging technologies at the United Nations. He said AI brings huge potential for vastly multiplied productivity and for large numbers of people to work in better ways, but warned that it also brings threats to human autonomy, via fraud and undermining identity, purpose and connection to society.
He said: “We want to make sure AI increases living connections, that we are not eliminated. That it makes a good society. The society where a number of people are excluded is not going to work. It will create problems.”
Elsewhere, attendees focused on how to gain business value from AI. This included Aiden Gomez, CEO of Canadian company Cohere, which specialises in use of LLMs in enterprises.
He said: “A generative model is kind of like a CPU – it’s a general piece of technology. You could deploy it inside any vertical for any purpose, like a CPU. But, in and of itself, just owning a CPU isn’t valuable. It’s what you build with it that is valuable.
“So, for that piece, you do have to be technical. You need to be a developer, to be able to build something on top of this model to create value on the other side.”