As a former neuroscientist, it's both amusing and sad to see AI people run into the same issues neuroscientists have grappled with for decades, as LLMs start to approach biological systems in complexity. (And mind you, LLMs are still smaller/simpler than the human brain.)

My hope is that the greater accessibility of neural networks over actual neurons will lead to some insights into biology. But I suspect they're still different enough, both in size, but more importantly in organization, that it won't be realized anytime soon.

The Golden Gate Claude is reminiscent of searches for the "grandmother" neuron. We know from direct brain stimulation that certain neurons are strongly associated with certain outputs...but we also know that a lot of knowledge representation is widely distributed. The whole thing is a mix of specialization and redundancy.

▲

nickm12 2 days ago | parent [-]

I was, for a time, a neuroscience major and have had this same thought. I'm concerned that we're treating these systems as engineered systems when they are closer to evolved or biological systems. They at at least different in that we can study them much more precisely than biological systems because their entire state is visible and we can run them on arbitrary inputs and measure responses.

	▲	SEGyges 2 days ago \| parent [-]
		I agree, what we do is much closer to growing them than to engineering them. We basically engineer the conditions for growth, and then check the results and try again. My best argument that insights from neuroscience will transfer to neural networks, and vice versa: For sufficiently complex phenomena (e.g., language), there should only be one reasonably efficient solution to the problem, and small variations on that solution. So there should be some reversible mapping between any two tractable solutions to the problem that is pretty close to lossless, provided both solutions actually solve the problem. And, yeah, the main advantage of neural networks is that they're white-box. You can also control your experiments in a way you can't in the real world.