I think using LLMs with RAG (aka tools) is more useful and reliable than relying only on training data that the model does its best to represent.

For example, using a search engine to find results for a query, downloading the first 10 results as text, and then having the LLM answer subsequent queries about those sources, or another example would be uploading a document and having the LLM answer queries about its contents.

This is also advantageous because much smaller and quicker models can be used while still producing accurate results (often with citations to the source).

This can even be self hosted with Open WebUI/ollama.

https://youtu.be/u3qJlWQNNxA

Sometimes being kind doesn’t necessarily mean being nice.

One of the first things I needed to do after creating a Lemmy instance was block all the AI bots from scraping everything. It only took a day before ChatGPT, AmazonBot and Alibaba came crawling through. Alibaba was the most annoying as it didn’t even identify itself properly, as mentioned in the article, it looks like an Edge browser.

They’re genuinely difficult to detect early on, however, systems to limit the damage they do, and ultimately push them out, are viable.

It could be viable to screen high impact positions through an assessment from a qualified psychologist specialising in Cluster B, however, probably not all positions.

Most people have zero chance of dealing with these individuals, even knowing what they are.

This comment seems ironic given the username.