What are some of the top risks around LLMs?
1. Hallucinations: LLMs sometimes make up facts and responses which are false.
2. Toxicity and bias: LLMs reflect the data they were trained on, so they may generate responses with toxic or biased content.
3. Job losses: More generally, many people worry that LLMs and AI will displace humans in some jobs, though similar historical breakthroughs indicate that jobs will likely change but not be lost overall.
4. Malicious AI: As LLMs and AI approach Artifical General Intelligence (AGI), some people fear that an AGI may become a threat to some or all humans.
5. Environmental impact: LLMs and big AI models require lots of energy to train, and therefore have big carbon footprints.