If chat GPT were anywhere as advanced as Data I would 100% accept him as a coworker. The problem with LLMs isn’t that they’re not people, the problem is that they lack any sense of logic and abstract reasoning, relying on heuristic analysis to cover that weakness. As a result, they’re quite frequently wrong, and don’t have a way to know when they might be wrong, leading to them presenting every answer with the same level of confidence.
If chat GPT were anywhere as advanced as Data I would 100% accept him as a coworker. The problem with LLMs isn’t that they’re not people, the problem is that they lack any sense of logic and abstract reasoning, relying on heuristic analysis to cover that weakness. As a result, they’re quite frequently wrong, and don’t have a way to know when they might be wrong, leading to them presenting every answer with the same level of confidence.
Remarkably like Baby Boomers
EXACTLY like Baby Boomers in government, in fact.
Chatgpt can use contractions