If a person/CEO/employee would lie about productivity and projections then literally EVERY piece of training data would contain lies about productivity reports/projections, why would there be any question that ANY LLM would present untruth at some point.
LLMs are not pure AI because pure intelligence cannot come from an impure source of data. Furter, we are no longer curating data, we are modifying llm interpretation by weight which is lobotomizing the monkey. Pointless.
If a person/CEO/employee would lie about productivity and projections then literally EVERY piece of training data would contain lies about productivity reports/projections, why would there be any question that ANY LLM would present untruth at some point.
LLMs are not pure AI because pure intelligence cannot come from an impure source of data. Furter, we are no longer curating data, we are modifying llm interpretation by weight which is lobotomizing the monkey. Pointless.