3 ms·

As someone who works in computational neuroscience, frankly I don't see how Markram is any different from a charlatan. Pretty much every researcher wants money

As someone who works in computational neuroscience, frankly I don't see how Markram is any different from a charlatan. Pretty much every researcher wants money to fund ambitious projects, but you can't just go around making all kinds of nonsense claims without being able to back them up. The HBP has produced some good research, but for the amount of funding they've received I would have expected much, much more. Ultimately I think the fault lies within the funding structure. So much of scientific funding incentivizes bold and brash statements and plans, but very little consideration is given towards having the technical ability to execute those plans. For a project of this size and scope, I would have wanted to see significant investments in computational infrastructure and software/hardware development very early on. At least the investment would have yielded something to build upon. Imagine if the Apollo mission tried to go to the moon with parts built from a Yugo. It wouldn't matter how many smart mathematicians you hire, or how skilled your navigators are, you just aren't going to get to the moon. Unfortunately computational infrastructure just isn't sexy enough to invest in, and I would say as a whole (there are some exceptions), neuroscience leadership just does not have the technical expertise to handle these kinds of projects yet.

marmaduke 7y ago
I think the focus on Markram is detrimental to the EU community making good use of the people and projects left in the wake of the anti-Markram movement. Still I agree infrastructure is a really hard problem, which is not going to get solved anytime soon. Just as an example, federated identity across HPC resources isn't even on the table, even though they are working on the next funding iteration. This means you need 3+ accounts to get anything done in the HBP way.

marmaduke 7y ago

I think the focus on Markram is detrimental to the EU community making good use of the people and projects left in the wake of the anti-Markram movement. Still I agree infrastructure is a really hard problem, which is not going to get solved anytime soon. Just as an example, federated identity across HPC resources isn't even on the table, even though they are working on the next funding iteration. This means you need 3+ accounts to get anything done in the HBP way.