Sign up to save your podcasts
Or
Share MyData Voices on AI
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
MyData Voices on AI
MyData Voices on AI episodes
- What makes us Human? In conversation with Harri Ketamo
Is AI learning the same way we do?
And if so, what is still uniquely human?
In this episode, Harri Ketamo explains why AI sees only a selected shadow of reality, and why the real question is who owns the data it learns from.
Harri Ketamo, PhD, has worked in learning sciences and artificial intelligence (AI) for more than 20 years, as both a scientist and an entrepreneur. He wrote his first game AI as a young boy in the late 1980s and went on to found Headai, a MyData Global member organisation that turns text into transparent, explainable knowledge graphs. He is an adjunct professor at Tampere University of Technology, a senior fellow at the University of Turku, and a 2017 Eisenhower Fellow.
In this conversation, we discuss:
Same mechanism, different starting point: people learn through years of observing and interacting with the world, while AI learns from material someone else has chosen for it.
Why AI is not the truth: every model reflects its training data and the rules and guardrails that control it.
When AI can decide: why a rule-based task such as taxation is very different from decisions that shape people’s lives, and why we should treat AI as a team member while staying responsible for the outcome.
Explainability in practice: why a career recommendation is only useful if you can see the reasoning behind it.
What our clicks reveal: how even the speed of our browsing says something about us.
Skills and art: what changes when machines take over tasks, and why Harri sees art as something that remains with people.
Who owns the data: what happens when AI learns our work from data we never chose to share, and were never paid for?
Chapters (approximate):
00:00 Introduction
02:16 From game AI to learning sciences
09:39 How people and machines learn
23:52 Judgement, choice, and trust
30:36 Why explainability matters
38:21 What happens to our skills and creativity
42:38 When should AI work for you, or with you?
Questions or ideas for future episodes? Write to us at [email protected].
46 min
About MyData Voices on AI
From the publisher's feed
AI now writes, recommends, and acts on our behalf. So what makes us human, and what do we want to keep deciding for ourselves?
MyData Voices on AI brings you in-depth…
conversations with the people who build AI and live with it every day, exploring AI, personal data, and what it means to be human.
Most of what we hear about artificial intelligence (AI) comes from headlines, forecasts, and product launches. We hear far less from the people who design these systems, use them in their work, or think carefully about their impact on our lives, in their own words.
Each episode takes one question and gives time to someone who has thought about it seriously. Our guests come from the MyData community: entrepreneurs, researchers, public and private organisations, and individuals working on how AI and personal data shape our lives.
Together, we explore questions such as:
What is still ours when machines can write, decide, and act for us?
Where should human judgement stay, and who is responsible when AI gets it wrong?
How can we trust systems we cannot see inside?
Who owns our data, and what happens when we no longer do?
How can AI work with people rather than simply for, or instead of, them?
The series comes from MyData Global, an international non-profit working for a world where people are empowered with their own data. Our work is guided by the MyData Declaration: people should be able to use their own data for their own purposes, have real agency through their rights, be protected from imbalances of data power, and understand and control how their data is used. These ideas sit in the background of every conversation.
New to MyData? Read the MyData Declaration and join the community at www.mydata.org.
If an episode made you think of someone, share it with them. Have a voice or a question we should hear next? Get in touch, [email protected]