Call us now:
The idea of using AI should be to reduce the time to conduct research and validate what is already known, and use reasoning to reduce costs, but also create value added.
But what are AI hallucinations? Hallucinations in AI are a misnomer, as AI has no consciousness; it lacks subjectivity, has no tacit knowledge based upon experience, is not self-aware, it responds to questions (direct or indirect inputs) and what it ingests based upon natural language. It is performing predictions based upon statistics. It just reached a different conclusion based on statistics.
There’s nothing new about the principle of technology-based hallucinations. GIGO still applies; you get what you ask for. Ask the wrong questions; the right questions in the wrong way; expect an LLM to have consumed and be trained on the data that you expect it to have, then you are asking for trouble. This applies when you use AI in any field. It becomes more dangerous if you give AI autonomy. AI is currently great at evaluating and automating already understood and structured data and processes. Not as great at working where there’s still a level of subjectivity required or reasoning across data sets that require nuanced, experienced, professional reasoning.
If you are a law firm and charge $2000 per hour, it’s reasonable to assume that you can afford to cross-reference AI outputs to validate outputs through expert and professional opinion.
@Thaddeus Dziekanowski Brian D. McCarthy
@Parava Security Solutions
#ai #aireasoning
https://lnkd.in/etu9qVQj
View Original Post
https://www.linkedin.com/feed/update/urn%3Ali%3Ashare%3A7452678627919499264