I had this analogy not long ago, insects might be 8 bit while we're 3D vulkan compute shader in 4K, but we still have the same game dynamics at the core.
I dunno about that one. I mean, a few months ago this "3d Vulkan" whatever blundered into the middle of a late-season fig tree also hosting a couple hundred bald-faced yellowjackets, which you'd normally expect to end badly except that the "8-bit" wasps recognized I meant no harm and opted not to interrupt their boozing to deal with me. Considering that every fruit still on the tree was quite literally crawling with burly black-and-ivory wasps - which, despite a lively interest in such animals, I totally failed to notice in my single-minded quest for bindweed roots - I think they might actually have been smarter than me.
I do like wasps, and I have some skill and maybe a little talent at being very close to them without eliciting dismay, but even I would hesitate to wander into a yellowjacket bar! Even so, their own skill, which I have often observed, at telling whether or not a human means them harm, no doubt saved me any number of stings. I even later the same day got some handsome macro shots of one of them who'd had too much and ended up crashing on my front porch windowsill to sleep it off: https://aaron-m.com/wp-content/uploads/2021/12/img_8240.jpg
jokes aside, i was simply making a dumb model based on brain size, in the sense that we are supposed to have more neurons than insects but maybe the core structure is similar, they just have less room for fat and shallow details.
That's fair, although it's also worth noting we share quite little in terms of neural architecture with even the smartest of insects, and our brains differ in neuron count by something like six orders of magnitude - I don't know of any insect with as many as one million neurons, most being around the 50-150k range, while human brains weigh in at around a hundred billion.
That said, to take this as prima facie evidence that insects are meat robots, the way behaviorists like to do, strikes me purely as motivated reasoning. There's no basis for it - no one yet has established an objective mapping between brain size or complexity and behavioral complexity, nor does this seem likely to happen soon. Ethologists, by contrast with behaviorists, spend most of their research time observing live, active animals in their own accustomed habitats - Jane Goodall, as opposed to that pervo freak Skinner - and it's no accident that all the really interesting results, from individual recognition by facial features in polistids to the likely range and spread of V. mandarinia in the Nearctic, are coming from biologists who take an ethological approach.
Well you know.. numbers. If you count lines of code on hard drive these days, it's also many order of magnitude more than old OSes. Some people made tiny lisp, tiny prolog that fits into a few kB while some use a full blown JVM class tree with god knows how many MB of code to achieve probably less.
Sorry but I don't understand. Does 'qualitative' stands here for 'something is missing, limiting what can be done with the device'? What can a FPU do which isn't doable using a CPU?
A CPU can do "anything" by definition (Turing Machine), but does it do it? :) You still need the function implementation to be present.
It is in terms of software, not on hardware - or, of hardware structures that perform functions. Sometimes, yes, «something is missing». Olfactory difference between man and bear could be possibly quantitative; control of the individual fingers is qualitative (a present or missing module).
