4 ms·
Are there any good AI models specifically designed for the "find all discrepancies/inconsistencies between X text and Y text" problem? It strikes me that this
by btown 4y ago
Are there any good AI models specifically designed for the "find all discrepancies/inconsistencies between X text and Y text" problem?
It strikes me that this could solve quite a few of ChatGPT's shortcomings by providing an automatic fact-checker - let ChatGPT create statistically-probable articles, then extract claims, generate search queries likely to find online reference articles from reputable sources for those claims, then compare the original claim against the text of each reference article, and escalate to a human if any inconsistencies are found.
Because it can fine-tune on specific reference resources for a specific generated text, it could prove more reliable than ChatGPT's gradual incorporation of this feedback as part of its adversarial training.
- james-revisoai 4y agoSometimes, documents can disagree with themselves; like a document detailing somebodies relationship history. Embedding time and when things change or making facts contingent on this can sort of avoid this, but most people imply things like time or common sense and it becomes tricky. Nonetheless there are hundreds of startups working on citable GPT and other AI approaches right now, and others working on semantic embedding technology that may be able to identify such gaps or inconsistencies. So it doesn't look too bad...
- btown 4y agoAre there any standout efforts here in your view?