crosen99 · HackerTrans

hace 4 añosdiscuss

In the short run, I’d think it more likely and useful for it to produce sources of factual statements rather than qualify or quantify its own confidence. For one thing, a source is something the user could verify. A confidence value would itself be another assertion the user would have to accept on faith.

crosen99

hace 4 añosdiscuss

It’s only a gut feeling, based largely on the predictions and accounts I’ve seen from those with actual first hand knowledge around the reasonable questions that you pose.

My framework for thinking about this at a technical level is that short term improvements can come from a number of factors, including optimizing the number of parameters, more data, better data preparation and selection, hyper parameter tuning, etc. I understand that many of the emergent capabilities we’ve seen recently from LLMs have come from more data, and I’m honestly not sure how to assess where short term accuracy improvement is most likely to come from.

crosen99

hace 4 añosdiscuss

Yes, it can be. But apart from the accuracy improvements I mentioned, which may make this a much less severe issue in fast order, there's also the notion that users will become more adept at navigating such perils, just as users developed skills for Googling effectively.

Consider, for example, how confidently doctors make assertions about critically important health issues only to be proved wrong by other doctors. Consumers know of this peril, and so have learned to ask for second opinions.

crosen99

hace 4 añosdiscuss

Those who object to or are skeptical about ChatGPT for search on the basis of accuracy issues or content rights are missing some crucial points.

Accuracy issues will improve. Mostly likely the improvement necessary for LLMs to answer a vast majority of queries as reliably as necessary will happen quickly. Not for all cases. But for a sufficiently large set of cases to make Google alternatives very compelling.

And content rights issues won't stop the march of chat interfaces anymore than they stopped the march of streaming services. What will happen is the evolution of new dynamics and rules in the means by which information is produced and information producers are incentivized and compensated.

The key with respect to content is that LLMs require source material to be produced in the first place. And, for the foreseeable future, AI will still rely on humans to produce that source content. Human labor forces will need to evolve and reorganize to optimize the production of original information in that setting.

comments