(I should say I'm a :LIB: and am reading the abridged Julian Borchardt version of Capital, so maybe there's an excised chapter that explains this or maybe there's a later chapter that does - I'm currently reading Decade of the Rate of Profit from Part II, chapters 13-15)
It seems that he insists surplus value is directly related to variable capital (i.e. labour), but other than semantically defining it that way, is this necessarily so? From the capitalist perspective how is labour really any different from any other input?
For instance, if one has a theoretical FALSC-style factory with no labour, surely one could still add surplus value to goods?
Over simplified example incoming.
So... think of all the bills that a company has to pay to keep functioning as a business. Think about all the bills/costs that the company has the power to negotiate with.
Can a company underpay the electric bill and still get electricity from a power plant? Not for long.
Can a company underpay its suppliers of raw materials? Sure, but not for long.
So where can a company underpay its "bill" but continue to be a business? Worker wages. If an employee doesn't like the pay, they can leave and another employee replaces them.
You talk about robots. In a horribly ham fisted way, a freezer case in a grocery store is just a big robot that holds and keeps product frozen. It does nothing else. But the company can't get the freezer case until it spends a "fixed" amount of money to get the equipment installed. Then a relatively "fixed" amount of money to keep the electricity to the freezer cases powered. There really isn't any way for the company to "pay the freezer case less" and keep the extra as profit as after the initial cost of installation and maintenance the company pretty much stops spending money on it until it breaks. (I know I know, freezer cases don't actually make anything but I think the sloppy example is simple enough to still be useful.)