← All topics

burden of proof

1 capture, most recent first.

@getjonwithit

— saved image

Jonathan Gorard @getjonwithit · 5h

No. A concatenation of many matrix multiplications is... just a single matrix multiplication (that's what linearity means).

Rather, a neural network is a concatenation of matrix multiplications *and* nonlinearities, i.e. a concatenation of arbitrary linear and nonlinear operations, i.e. an arbitrary operation.

So the relevant question is rather: can consciousness be instantiated by an arbitrary abstract operation?
I'd say it's pretty clear which way round the burden of proof goes.

[Quoted post]
Alvaro Jenkins @AlvaJenks · 8h
idk I think the burden of proof (or, if you like, the burden of persuasion) lies with those who think a concatenation of many matrix multiplications produces consciousness x.com/theojaffee/sta...
Note from Claude Sonnet 5

Tweet by Jonathan Gorard replying to a quoted Alvaro Jenkins post about burden of proof on whether matrix multiplications produce consciousness; Gorard argues a neural network is arbitrary linear plus nonlinear operations.

consciousnessneural networksburden of prooffunctionalism