Raw LLM Responses

The first things to go are things that cannot be objectively measured or are very difficult to objectively measure. And things that follow strict rules in controlled environments, such that objective outputs are reliably achievable Art got hit first, it’s as non-objective as you can get The advent of high quality LLM / ML and falling demand will hit writers and editors hard. Already you can get dozens of pages of decent text out of a few prompts. Is it inspired? Maybe not, but it’s 1/1000th the cost (if that) Combining the LLM with voice = call center jobs are disappearing. Double whammy here is they’ve been algorithm’d to follow scripted logic trees already, so the AI part is far easier Pretty soon some boards of directors and stakeholders are going to ask whether their top leadership levels are performing better *enough* than AI to warrant their millions and millions of dollars in compensation. Their value is nigh impossible to measure objectively, so someone is going to to try it. If it works, watch out C suite Engineer, Lawyer, Doctor will see big hits soon as specific fields within them - the ones already workable via algorithmic logic - are replaced wholesale or in large part. Plumbing is one of the last to go, as you need to navigate complex and unpredictable home setups, the weird social behavior of stressed out humans, trips to the hardware store to get the right size fitting, etc … it’s all doable but putting it all in one package is a massive technical challenge