The Future of AI is Collaborative

«AI will not pass the Turing test» – this was the bold prediction made by Microsoft Technical Fellow and Chief Scientist Jamie Teevan during her keynote.

Digital evolution of collaboration during the last decades has transformed the way we work together – including:

• Shift to cloud
• Remote and hybrid work
• Rise of AI with tools like Copilot

Nowadays, new forms of collaboration consist of writing a prompt, providing grounding context and passing judgement to the output.

AI is different from human intelligence. Computers are deterministic, while AI is probabilistic. This means that AI can generate unexpected results, which can be both a strength and a weakness. Like humans, AI can make mistakes.

These are four areas of differentiation which can be interesting for future AI development:

• Feedbacks: humans and AI can learn from each other; Language models are «happy» to make mistakes if they get feedbacks
• Speed: We can use AI to speed up our work, but we need to be careful not to sacrifice quality for speed
• Repeatability: what can AI suggestions look like from different perspectives?
• Scale: AI can help us to scale our work

When we take these different aspects and combine them with the facts that we know, we can create a new form of collaboration that is more effective and efficient than ever before: think of the interesting conversations we can have, think of all the knowledge we can share, think of the creativity we can unleash.