The Facts Behind DeepMinds AlphaFold’s CASP14 Breakthrough Announcement

DeepMind breakthrough with AlphaFold aids drug research ...

Deep Learning is a pattern matching algorithm. Quantum Computing is an ab initio method. it starts from scratch and is supposed to solve any problem of a specific type thrown at it. Congratulations to the DeepMind team. But apart from the vendor speak and jargons this is the announcement from the CASP14 organizers

"During the latest round of the challenge, DeepMind’s AlphaFold program has determined the shape of around two thirds of the proteins with accuracy comparable to laboratory experiments*. AlphaFold’s accuracy with most of the other proteins was also high, though not quite at that level."

Automatski believes the solutions to the worlds most pressing problems cannot be done using function approximation or pattern matching which is Deep Learning in entirety. We simply cannot train on known structures and try to construct new solutions from existing patterns. Thats NOT how we solve problems in a space of gazillions of unknowns. We need an Ab initio approach that can solve any problem thrown at it, amongst gazillions of possible problems at the scale of this universe. That and only that will define the next millennium of mankind. – Amen!

Here is the press release from CASP14 organizers https://www.predictioncenter.org/casp14/doc/CASP14_press_release.html

And here is the video from the DeepMind Team

Its a free world. Please believe whatever you want to believe. Automatski is only in the business of dealing with facts and millennium inventions.

Think about it. 😊

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: