Someone’s got to be able fight back against the robo-apocalypse

We might need to agree to disagree on that because I can’t see how it gets us any closer (as fascinating as it is)

“Because it just does” is exactly the point I was making.

No one is arguing that we haven’t made massive strides in being able to describe our physical environment in elegant mathematics, but just that it doesn’t get you any closer to explaining away the mystery of existence

Can’t you see that these gaps will never cease to exist? We could never reach the point of, “ah we have now explained away every possible scientific question” even with perfect scientists and infinite time

They are amazing discoveries but why did the Big Bang happen?

No one is claiming that we don’t know more than Chimpanzees, just that fundamentally we are never going to know why we/everything exists, because that any explanation would itself need explaining

But isn’t that what Ian is saying, that there will always be an unanswered question (eg why the physical laws are the way they are), not because we haven’t worked it out yet, but in if you came up with an answer you would be left with yet more questions (ie why do those laws exist?)

Isn’t the issue that to “explain” something (in my understanding of the word), there must be a more fundamental framework in which to describe it. Therefore, to be able to explain everything you would need infinite frameworks to describe the previous ones. “Turtles all the way down” if you will

I haven’t watched the talk and I would hope that he adds a lot of qualifiers to that slide, but even including it is problematic. I would agree that new advances are needed, but given what many organisms can do with far less units, I’d say that whether larger scale is is not at all certain

When you are dealing highly nonlinear systems, even a few % of variance in your base unit is likely to give you completely different behaviour at scale, let alone 30%. There are also many subcortical neurons an ANN node wouldn’t explain any of their behaviour as they don’t integrate and fire

I’m interpreting “reasonable abstraction” in the context of the presentation in the OP, which assumes that it is close enough that you can expect the behaviour the brain to simply pop out by scaling up the ANN. This was a ridiculous assumption when the Blue Brain tried it even with HH models.

That's great you have used multi-compartment models but you are not showing much evidence of "knowing these things". The claim that ANN nodes are a reasonable abstraction of all the different types of neurons in the brain is just not "factually true" because they don't spike intrinsically

My man, you need to explore a bit deeper across the interface into biology. Even with the Izhikevich model you can get intrinsically spiking neurons

What are the protections against bots?

Putting aside the age restrictions for a moment, it could actually be kind of cool if you could design a service that allowed you to authenticate yourself as a real person and generate an anonymous tokens to access social media, as a way to combat bot accounts.

I presume you would only need it to sign up too

How are they going to know that an account is Australian?

My family left SA when I was 3, and one of my only memories of that time Golden North honey ice cream we had as a Friday treat. Used to come in a roll like devon (or fritz for the locals). Every time I go back for a visit I make sure to track some down, so good!

Big batteries and EVs to the rescue again as faults with new nuclear plant cause chaos on Nordic grids Europe’s newest and most powerful nuclear reactor – delivered more than a decade late and nearly four times over budget – is proving a headache for grid operators now it’s finally up an…

Europe's newest, most powerful nuclear reactor, delivered over a decade late and nearly 4x over budget, is now causing headaches for grid operators. Big batteries and EVs stepping in to stabilise Nordic grids amidst the chaos. #NuclearPower #renewables #auspol

reneweconomy.com.au/big-batterie...

Research funding for both discovery & medical science in Australia (ARC + NHMRC) has been static for over a decade (at about $800 million each). Wages have grown by 60% in this time.

Why are both sides of government defunding research?

He’s got to be just trolling at this point

James Patterson is definitely an anti-hero

Nice, made a nice soundtrack to my train home yesterday

What are you listening to?

Hellooo Bluesky! I am trying to grow my tiny lab at Vanderbilt and am recruiting 1-2 postdocs who will help me tackle addiction heterogeneity. Please reach out if interested! All reposts are appreciated! www.vanderbilt.edu/postdoc/posi...