Latest AI worldview from the GAIN conference in Riyadh
Charles Cadbury: 16th September 2024
A week working in Riyadh has opened my eyes to a maturing view on what’s going on in #AI from a multidisciplinary perspective and the future has never been clearer.
I’ll try and distil some of the thinking here and I’m reaching out to our partners to explain how this plays into our roadmap at Say It Now and broader planning for 2025 and beyond.
SETTING THE SCENE:
I was invited as one of 200 speakers to the global AI conference in Saudi Arabia, focused on an ambitious 2030 vision accelerating the region away from reliance on oil towards leadership in technology, tourism and sustainable living.
This is one of the largest AI conferences in the world, held in the King Abdulaziz International Convention Centre under the themes ‘Now, Next, Never’ covering off the current state of the art, what’s upcoming and exploring the ethical considerations of developing humanity shifting technologies.
THE RANGE OF EXPERTISE
I was lucky enough to meet and have in depth conversations with pioneers setting the vision and applying AI in the fields of mental health, genomics, Art, Venture, Marketing, Creator Economy, Neuroscience, Journalism, Government AI Development, Education and plenty more talks and incidental conversations across the week. Being immersed within these differing perspectives as opposed to the industry specific conferences I often find myself at has offered a broader understanding of the broad applications of the new technologies being developed arounds us.
COMMON THEMES AND OPPORTUNITIES
We’re in execution mode. When chat-GPT launched there was a period of wonder and pontification about what this type of technology meant to society and where the opportunities lay. Whilst this narrative continues today, real applications have been developed, launched and real-world feedback is flooding in, allowing the data required to create the feedback-loop flywheel which accelerates progress.
Many of the most obvious applications are based on LLM’s. I met a lot of people who had ‘built a proprietary LLM for X’, presenting advice, streamlining processes or acting as valuable co-pilot in a specific domain. This is where I believe a huge amount of opportunity lies, at some point every specific task or expert domain will need its own agent. In Say It Now’s domain that’s an advertising campaign agent.
Streamlining processes in healthcare, as it can touch so many people is abound, from drug discovery to patient testing, interpretation and processing. Some states are keen to sequence their entire population to inform how they set up their national healthcare provision. The scale of ambition and targets set to deliver these are inspiring and even more incredible now, seem to be within reach.
SUMMARY / REFLECTION
A common language for rapid dialogue between practitioners is emerging allowing for product vision and application to be better understood, tools are maturing, consumer appetite and acceptance is rising steadily and there is a huge amount of ambition and optimism for what is to come.
There is a lot in market and the key challenge is being where your audience is to understand how to engage in new ways, understand what’s working and iterate. This has always been software development best practice but we can now do this faster than before, delivering more material improvements. There has never been a better time to evolve how you engage with and serve your customers.
…and finally. I met a retired WSJ journalist and now AI podcaster supreme, Craig S. Smith, along the way who had covered the region 20 years ago, had come directly from ‘The Line’ in NEOM and was able to put some real texture to the scale and pace of change happening in The Kingdom of Saudi Arabia. A real highlight was seeing how dextrously he was able to pull the real mood of the city from some locals voices we spoke to over a sheesha. The mood of the youth is incredibly positive and seems to be aligned with the country’s 2030 vision. The pace is real, check out this video for a flavour: