When OpenAI released ChatGPT in late November last year, they added even more fuel to the AI hype. Test users joined at a record pace, and two months later, ChatGPT is estimated to have reached 100 million active users. Are we part of a historic moment with a before and after?
In "Future of E-Commerce," Björn Påhlman Spenger meets entrepreneurs and specialists in e-commerce to inspire others with tips, advice, and exciting stories. Björn invited Martin Rosvall, co-founder and lead scientist at Sift Lab, to talk about AI and the limitations and opportunities of machine learning in e-commerce and retail.
As a professor at Umeå University, Martin, along with his interdisciplinary research group, develops new algorithms to simplify and highlight important structures in large data sets. The goal is to explain various phenomena in complex systems and predict how they will evolve. This includes everything from understanding why some are severely affected by Covid-19 while others have mild cases, to the consequences of the changing distribution of species on Earth. The systems may seem different, but the key to understanding them is the same: How does the pattern of interactions affect their function? In e-commerce, interactions are about who buys what and when, and methods developed to predict which proteins will interact, what the next word in a sentence is, or which product a person is interested in right now share many similarities.
To look into the future and see where we will take AI and where AI will take us, Björn and Martin first looked back. First to the Turing test, devised by Alan Turing in 1950 to answer the question of whether machines can think. Then to the development towards general intelligence and the singularity - and when we might reach it. Further to ChatGPT and the future of e-commerce.
However, the development has not been straightforward. Since the 1950s, several AI summers with inflated expectations and the belief that general intelligence is just a decade away have been followed by stagnated development and withdrawn investments. Even though self-driving cars are not yet the norm, which was predicted as recently as five years ago, we are currently experiencing the longest AI summer thanks to the ongoing digitalization with access to cheap storage and computational power.
Sift Lab is one of the many suns of this long AI summer, and you can listen to Martin's vision of the data-centric computational world of the future in episode 146 of "Future of E-Commerce."This text was written by a human.
You can listen to the episode with Martin and Björn here:
https://podcasts.apple.com/se/podcast/framtidens-e-handel/id1536993379?i=1000597807352
