"We don't have the ability to [...] scale our intelligence horizontally (e.g. by copying it to more hardware), which an entity executing on a computer system could have. That's one of the main points of the argument for singularity."

Count me as skeptical about that. Scaling horizontally means making it a communications problem, which the last 50 years have hinted is a bigger deal than you might think.

> Thats a very strong claim which you provide no basis for, in that comment or in your follow-ups. Maybe replication at the level of neurons, as opposed to atoms, would work? Or maybe at an even higher level? We don't know yet.

Actually, worded correctly, it's a reasonable claim (assuming pure materialism). If there is nothing but matter and energy and their interaction, then a perfect simulation of all of your cells would undeniably be a perfect simulation of you.

The strong and unsubstantiated claim for which no basis is provided is that simply replacing the neurons in your brain would be "enough." There are good reasons to believe that the neurons in your skull are not sufficient (are you still you absent sensory input?) and apart from a lot of philosophical navel-gazing there's no proof forthcoming. Even pure materialism is an assumption--a good and reasonable one when it comes to doing science, but not something I'm willing to gamble my consciousness on.

Of course we can scale our intelligence horizontally. We have almost seven billion of us walking around. It turns out scaling horizontally gets you a lot, too. You might be able to scale horizontally more efficiently if you had a bunch of brains-in-vats that are telepathically linked into some sort of hive mind, but probably only by constant factors.

We can also self-modify within certain limited parameters (learning, drugs, etc.), but that's the most any intelligence can do. Whether we can widen those parameters by more than a constant factor is an open question.