Kathy Gibson reports from Gitex – In 1996, chess master Garry Kasparov agreed to a match against IBM’s Big Blue computer – and lost.
In analysing the game, Kasparov came to the conclusion that he hadn’t in fact been beaten by a machine but by a human, because it was a human that designed the machine’s processes.
Dr Anish Agarwal, head of global analytics, thinks this is an important point to bear in mind as we navigate the world of AI: that there is always a human behind the machine, even as we get closer to forming a relationship with technology.
Describes AI as a collective term for the capabilities shown by learning systems that are perceived by humans as representing intelligence. And generative AI (GenAI) is the latest evolution in AI, characterised by its ability to generate novel content based on pre-trained models.
The stages we have passed through on the road to this point are analysis, then prediction, then deep learning and now GenAI.
Throughout the journey, data and AI have become critical driver of total enterprise re-invention, Dr Agarwal tells the AI Anywhere conference at Gitex Global 2023 in Dubai.
“AI enables insights at speed, and can improve the outcome of multiple business process, helping to make them more efficient.
“In addition, it can be good for the bottom line, reducing costs and increasing revenue.
“And it helps to reduce risk, helping companies to become more compliant, and enhancing their audit and risk initiatives.
“Finally, AI can result in improved experiences when it’s embedded into engineering, operations and workflows to improve customer experience (CX).
At the same time, AI has the potential to reshape society in realtime, Dr Agarwal adds, so we need to do more to ensure it is developed in a human-centred way.
Design thinking principles can be used to ensure AI is focussed on solving the real-world challenges of everyday people, he says. “It also needs to help people answer what went wrong, and give them the tools to explore the art of the possible.”
Dr Agarwal believes that a human-centred approach to AI needs to embrace four key tenets:
* Develop AI apps for the common good and benefit of the community, co-created with and deployed by people – for and with humans.
* Equity at the centre – design AI apps to be accountable to humans and controlled for bias, particularly for his marginalised, the disabled and highly sensitive populations.
* Secure – build AI apps with end-to-end privacy and consent for how data is used.
* Transparency and accountability -be clear on when AI is being used, how it works and who is responsible, so you can create trust in the experiences it creates.
He points out that the Institute of Electrical and Electronic Engineers (IEEE) has released its general principles of ethically-aligned design for intelligent systems.
These principles embrace human rights, wellbeing, data agency, effectiveness , transparency, accountability, awareness of misuse and competence.
In addition, Dr Agarwal says its important for people and organisations to understand the AI development framework from ideation to design to production.
“We usually focus on production and outcome, and tend to forget ideation or to make sure it’s inclusive. But if you bake these principles into the design it will solve a lot of problems.
In addition, he adds the quality of the data used by AI systems is vital. “If it’s not good quality, the effectiveness of the systems will be poor.”
And – something that should be obvious but is often not done – Dr Agarwal urges developers to test their systems before going into production to ensure the effectiveness of the algorithm and the platform.