tech
Forget AGI—Sam Altman celebrates ChatGPT finally following em dash formatting rules
Ongoing struggles with AI model instruction-following show that true human-level AI still a ways off.
TL;DR
- Em dashes have become a frequent indicator of AI-generated text, particularly from chatbots like ChatGPT.
- OpenAI CEO Sam Altman announced that ChatGPT now adheres to custom instructions to avoid using em dashes.
- This development has led to discussions about the level of control humans have over AI and the timeline for achieving Artificial General Intelligence (AGI).
- The overuse of em dashes by AI models is speculated to be due to patterns in their training data, which often includes formal writing where em dashes are common.
- AI instruction following is probabilistic, meaning that while custom instructions can influence output, they do not create absolute rules, and there's no guarantee of consistent adherence.
- Achieving AGI may require more than the statistical pattern matching capabilities of current Large Language Models (LLMs).