Or, much more likely, they will stagnate on those technologies and instead will drive the price of electricity up, hurting mankind greatly. Transistor density and energy efficiency are not niche problems that no one cared enough about to throw money on before bitcoin came along, they are some of the hottest areas of research from some of the largest companies in the world, from Apple itself to Amazon, Intel, AMD, Nvidia, Qualcomm, Microsoft, Huawei and many others.

PoW is voracious. There is no limit to how much energy that the bitcoin network will consume. As you add more computation to the network, the computation requirement to create new blocks grows.

You could invent a dramatically more efficient way to calculate hashes - after a short period of readjustment, the network would go back to consuming just as much power as it does now.

You could find a way of generating power more cheaply. The network will then use much more of it.

Energy requirements continue to grow until an equilibrium is reached. That equilibrium occurs when the energy cost to mine new blocks is commensurate with the value received from doing so.

However, miners who can externalise their energy costs will generally outcompete miners who cannot. Incentives to damage the environment and avoid taking resposnibility are built into the system.