Bullshit jobs did not arise because people thought they were actually useful. AI can not replace or reduce bullshit jobs, because *the point* of bullshit jobs is that they do not have any impact. They are clientele politics and make work schemes, either from the government, zealous consulting companies or from the management themselves.
AI does not make bullshit jobs obsolete. AI makes it so that bullshit jobs become more and more necessary and will be forced upon more and more companies.
We built our own layer shared across the team. Like openclaw, but with sane defaults that already has the setup for security and cost and integrations in place. See: norisessions.com
Bullshit jobs did not arise because people thought they were actually useful. AI can not replace or reduce bullshit jobs, because *the point* of bullshit jobs is that they do not have any impact. They are clientele politics and make work schemes, either from the government, zealous consulting companies or from the management themselves.
AI does not make bullshit jobs obsolete. AI makes it so that bullshit jobs become more and more necessary and will be forced upon more and more companies.
Nice post - thanks for sharing the examples, super helpful! What's the gist of the how here? OpenClaw?
We built our own layer shared across the team. Like openclaw, but with sane defaults that already has the setup for security and cost and integrations in place. See: norisessions.com
A lot of software engineering pre agents was deciding what to build because writing software took a long time.
With agents making new code quickly, what new constraints if any is your team running into?
Verification debt, by far. We wrote about this here: https://theahura.substack.com/p/agentics-cost-to-implement-vs-cost?utm_source=share&utm_medium=android&r=5sutf