@xtaldave @graemewinter @markus I'd like for the source code to be available as well, but I do understand that they are now turning it into a commercial product.
I feel that even if they now block academic access to AF3, we should still thank them. They published AF2 which was a significant step forward as well as providing the colab resources for free.
@graemewinter @xtaldave @markus Yes, they are doing it to make money. But I have used extensively AF2 and I feel they did greatly help me in my research. I believe they gave back to the community AF2 which is already a huge thing, as well as the idea that it can be improved.
They are letting academics make limited use of AF3.
I don't feel like they misbehaved in any way.
Sure, they could have just considered AF an academic project and kept it fully public.
I don't know, I see a bunch of IT companies now trying to move part of their business towards drug design.
Google, nvidia, microsoft.
I can understand they want to use it as some competitive advantage over the others.
I still believe this will be beneficial to the community in general.
The Nature article is an advertisement, that is clear. Probably should not be on Nature itself for what it is worth. But on the other side, we may give a little leeway to the people who did bring great innovations to the field.
@rastinza @graemewinter @xtaldave Telling us it's possible, but not how, is not innovation, and doesn't benefit science or society.
@rastinza @xtaldave @markus they trained the system on data which the community have made open over more than half a century
So, sure, they can make it closed and commercial and all, but don't expect me to toot about how awesome science they are: they are doing it to make money