Harnessing Frontier Technology

This was an awesome panel that I just hosted at The Digital Economist at #Davos with 

➤ Joseph Raczynski, CEO & Futurist, JT Consulting & Media

➤ Dr Dimitrios Salampasis (PhD, F-FIN, MAICD), Senior Lecturer- Emerging Tech & FinTech, Swinburne University of Technology

➤ Bhuva Shakti, Board Advisor & Chief Sustainable Innovation Officer, Wallet Max

We discussed frontier technologies and AI took the top spot with associated concepts like agentic AI, blockchain, quantum computing among other things. Joe spoke about his experience with Agentic AI and how the third evolution of Chat GPT was to train AI to be an agent for you and now the next evolution was to have AI agents work with other agents. He stressed this was both exciting and worrisome as there was so much potential but so much that could go wrong. Gartner estimated that by 2028, 33% of AI will be agentic AI. This topic has been dominating conversations given by Big Tech so something to definately watch out.

These types of technologies require huge amounts of data which has these complex ethical challenges. We are at the point where more data is being made by machines that feed the AI models. Some simulations show that when the synthetic data or machine or AI-generated data exceeds human data, the AI models collapse. So data alone is not enough – you need digital infrastructure and you need smart minds managing these powerful frontier technologies. Dimitrios addressed the challenges with open-source innovation and the narratives that assumed they were only beneficial. The challenge he highlighted was to understand what was open, why it was open and who was managing and regulating it. Not everyone who is part of the community may have interests for good of all (just saying - Git Hub is now part of Microsoft). So while the intension of open innovation was collaboration, we needed more understanding on why this was good and needed.

If we wanted responsible fronter technology – the question came up - how do we build it for the good of all when women are excluded from the development of AI? Bhuva highlighted that the funding for women founders working with AI had decreased post-pandemic (not that it was very high or equal to that devoted to men). And women are 50% of the global populations! How can we get more women in the conversation, work on areas like climate tech that are critical for future generations and have a more caretaking attitude to people and planet – not just profits. She highlighted the need for responsible entrepreneurship. The challenge is that we do not have enough data on women representation and achievement in this space to make meaningful high impact decisions. 

This data paucity is worrisome as frontier technologies like agentic AI or quantum computing are dependent on data volumes. Just because you had data did not mean that it was not biased or representative or relevant to the problem being solved. Dimitrios highlighted this narrative with the example of El Salvador and bitcoin. El Salvador adopted bitcoin in 2021. It became a pioneering test-case. What the media narrative misses is that it did not lead to financial inclusion due to poor wide-spread adoption. The algorithm bias in our online information feeds is proving to be challenging as often policy makers are working with incorrect narratives (in another panel Ann Dunkin, Chief Information Officer, United States Department of Energy reiterated the need for strong good journalism).

This conversation brought us to the importance of AI literacy. Joe highlighted that the technologies like quantum computing could not just do great good but great harm, and it would be difficult to control for this. The way we were using AI across sectors meant we could even have hybrid humans in the next few years (Japan has just given permission for research combining animal and human DNA). This adjacent possible into new industries leaves us with unexplored governance areas where the only precedent is science fiction stories.

Bhuva reiterated the need for responsible investing and responsible entrepreneurship. Without this we would be unable to create a world were frontier tech was used for good. We ended the panel acknowledging that AI as it was, was not inclusive and the current IP law did not “fit” into the problems of the future. We left the attendees with more questions than answers but – this is what we do when we are working on creating a safer world for tomorrow - open up the dialogue for more discussion, so together we can thoughtfully craft a world where all of humanity and the planet can flourish in – not just the the current few.