Daily Junction Play
718watching now
The Royal Institution β€’12 Jun 2026 30 here now

AI is solving science's biggest problems β€” but has no idea why it's right | with Claire Malone

What does it mean to make a scientific discovery β€” and can a machine really do it?
πŸ“… Filmed at the Ri on 2 May 2026
AI is now simulating particle collisions at CERN, predicting protein structures that

For You

Because of what you watch

Comments (20)

Daily Junction discussion mixed with clearly attributed comments from the original video provider.

Join in β€” free. Comments on Daily Junction are for members, so real names stay rare and bots stay out.

One field. We email you a 6-digit code β€” no password needed. Your comment is kept while you do it.

Under 13? You’ll need a parent’s OK first β€” it takes them one click.

@
"is science about finding patterns that work or about understanding why they work?" β˜„πŸŒ πŸŒŸπŸŒž
@
πŸ’— I feel like my brain is making popcorn when I listen to Claire's presentation and I love it
@
Great analysis
@
Thank you :face-blue-smiling: learnt so much
@
Thanks for this presentation!
@
Soy millonario y estoy tomando mate desde un puesto de guardia de seguridad en el puerto de Montevideo πŸ‡ΊπŸ‡ΎβŒ›β³,millonario porque tengo una familia y porque puedo comprender absolutamente todo lo explicado por esta gran cientΓ­fica πŸ‘ rico de poder estar presenciando el principio y el fin de un momento dorado de la creatividad humana. A principios de siglo xx con el cine ,la mΓΊsica y la ciencia y hoy con el final del ingenio humano..son muy pocas las historias que quedan por contar, las hemos contado todas, desde ahora solo vamos a querer borrar memoria para poder vivir nuevamente como si nunca lo hubiΓ©ramos sabido..muere de a poco los hΓ‘bitos,la familia, los rituales..nacen todas las respuestas a todo..
QuizΓ‘s la muerte sea un privilegio en este punto..prefiero morir que vivir eternamente en una computadora..es como poner un cd de Mario bros y que Mario busque cada vez y en cada inicio la misma meta y con el mismo entusiasmo pasar cada pantalla..
@
@3:43 thank you
@
I love Artificial Intelligence. It's an incredibly useful tool.
@
Thank you, Claire!
@
AI is currently like the first ENIAC. In some time, 5, 10 or 20 years from now, it may approach perfection. Continuing is necessary.
@
Love you
@
@
Wow, girl!!!!!!!!! Fantastic!!!!
@
What an amazing woman! Such an inspiration and a excellent lecture ( packed with information) zβ€β€πŸŽ‰πŸŽ‰πŸŽ‰
@
Excellent talk, thank you very much Claire :)
@
Great tips, thanks for sharing!
@
Excellent and inspiring in every way!
@
38:06 Smoothing the curve for aesthetic appeal is the most AI thing ever :D
@
Thank you for useful information!
@
πŸ’¬ Claire ends this talk with a question from statistician George Box: "All models are wrong, but some are useful."
AI is already finding protein structures, spotting anomalies at CERN, and designing experiments faster than any human researcher. But as Claire argues, it still can't tell you why a result is true, or which discovery is actually worth pursuing.
So where do you stand?
Let us know below πŸ‘‡