Is AI destroying our Institutions.

A team at Boston University has given the blueprint for how AI is destroying our institutions.

They argue that AI eats away at the human processes that institutions are built on. That it drives

→ cognitive offloading: people stop thinking deeply (automation bias)
→ skill atrophy: people stop practising judgment (learned helplessness)
→ "workslop": Cleaning up AI-generated errors kills productivity (social proof)
→ false confidence: the illusion of accuracy suggests expertise (degradation of mastery)

Not only that, AI short circuits the friction that makes for better decision making. It makes it 'too' easy. When you build muscles, you have to progressively overload and break your existing muscles, creating scar tisue which strengthens and grows the muscle. When AI removes the friction, your decision making muscle weakens and dies.

The ease of using AI leads to less reliance on social networks and trust building, which over the longer term impacts the depth of relationships that are built, the mycelium that keeps an organisation alive. Reciprocity reduces, people will become disengaged and become transactional. The emotional conviction in an organisation, a team and their purpose erodes, leaving a hollow shell.

In many cases the AI is a poorly trained intern, that is still turning out 'slop', which takes oversight and management. The training will get better and it will have better context of why choices are made, but until then it will not be seen as a fully contributing member of the tribe. The trust will need to be earned.

The biggest argument is one of the strongest for me.

Many of these factors are not new and are not unique to AI. AI shows the weaknesses that institutions already have and accelerates them.

The stretched capacity, short term focus, performance pressures, poor leadership and erosion of trust. These are weaknesses now.

Ironically, AI could probably help you spot (and fix) the weaknesses that it is exposing in your organisation.

SOURCE
Paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5870623

BESCI AI OPINION

This paper raised more questions than it answered for me.

→ The role of institutions is to manage people and their work effort, do we still need them?
→ How quickly will our organisational weaknesses become our death spiral?
→ Is AI really only as good as the past and what it knows, or can it be taught, like a human to be innovative and think of the future.
→ The ease and reliance of using AI, which is sycophantic by design accelerates low diversity, where you surround yourself with people who think like you. It doesn't have to, but friction will reduce usage.

However much you teach a robot or AI, it will never 'feel'.

It may have the ethical guidelines, but it doesn't have that twitchy feeling that something is off, or the stomach turning anxiety that something is wrong. These are what have served us well. When that responsibility is given to just a few humans who oversee, is that too much.

A complaint is a gift ... by knowing your weaknesses you can do something about them, if you have a flexible, agile, coherent organisation, or is that just another weakness?