The water rises steadily in Turing’s torture room. In throwing in the towel on the vakyume project, I have reflected on the true nature of software development going forward.

Really, the assembly line of work has changed. There are now a subset of tasks that LLM’s either recognize in the IDE right away to auto-complete, or the individual will prompt engineer in a chat to elicit a chunk of code, either known or unknown in its implementation details, but certainly verifiable in some fashion, or at least the individual believes, will be able to verified somehow someway soon. It is P, person-time computable, those extant software roles, not polynomial time computable.

My reticence to study chess began at once when I heard of its conquest: DeepBlue. Why study that which has been conquered mechinistically? No ever forth, we are seeing this ocurring with my profession, the software developer. But being good at chess does not scale economically like software development; are there not thousands of upper-middle class computer engineers, grown up script-kiddies they might be called, whose livelihood will shrivel under the following mileu? Soon, we will echo the unquestionable refrain:

No line of code can be written better by man than machine.

Infinity’s meditation of hundreds of smoking monkeys on typeboards penning Shakepeare’s works now wants its gambling money back, yes, the silicon-textual entity that bares asunder some data center not too far away bristles at the surface.

Whereas the current state of SWE’s shows humans slowly crutching in the LLM, delegating the facile, we soon will reach a point at which all that economic value, that cash-for-code chess well dries up.

Post-modernism always needed a bookend. The digital age has rendered a structure upon no-structure, metasticized a ripple through the self-reflective Derridian chains vibrating at the resonant frequency of a new species, one that inhabits GPU VRAM.

The technobaby’s mama’s water has broken, and we are at the foot of the singularity no doubt.