i just recently heard its use pervasively in some dental offices since 2024, and they often give hallucinated areas on a tooth leading to unneccesary procedures and costs, this is probably why it wont be diagnosing cancer anytime soon(because it will likely hallucinate a tumor more than not.
Diagnosing is a separate process that involves several steps, so I’m setting that word aside but
Identifying cancer in imaging or pathology smears is one of the better things certain types (not generative, not LLM) of AI is actually good for. But again it’s specifically tweaked programs that aren’t just general usage.
Leasing to unnecessary procedures and costs is probably the reason why dental offices would be interested, lol.
Though to clarify, did dental offices started using LLMs?
The things is that LLM and to a lesser degree other generative AI (generating pictures, music, videos) are the things that generated current hype and they have the hallucination issue.
The traditional AI that we had for years like image recognition, or solving specific problems (for example driving a car), machine learning (back then we were more humble and that name was used to emphasize that it doesn’t actually think) etc those things are as they used to be and they don’t have this specific hallucination issue. The hallucination is really problem of the LLMs, but looks like now people call everything AI, which technically is not wrong (even things that we take for granted like GPS navigation or speech recognition are AI), but that help LLM obscure itself with other intentions.
I actually hope dentists use the old proven image recognition not asking chatgpt which is a chat bot that also incorporates other technologies to make it look better than it actually is.
i just recently heard its use pervasively in some dental offices since 2024, and they often give hallucinated areas on a tooth leading to unneccesary procedures and costs, this is probably why it wont be diagnosing cancer anytime soon(because it will likely hallucinate a tumor more than not.
Diagnosing is a separate process that involves several steps, so I’m setting that word aside but
Identifying cancer in imaging or pathology smears is one of the better things certain types (not generative, not LLM) of AI is actually good for. But again it’s specifically tweaked programs that aren’t just general usage.
https://futurism.com/the-byte/ai-pastries-cancers
https://pubmed.ncbi.nlm.nih.gov/31110349/
Leasing to unnecessary procedures and costs is probably the reason why dental offices would be interested, lol.
Though to clarify, did dental offices started using LLMs?
The things is that LLM and to a lesser degree other generative AI (generating pictures, music, videos) are the things that generated current hype and they have the hallucination issue.
The traditional AI that we had for years like image recognition, or solving specific problems (for example driving a car), machine learning (back then we were more humble and that name was used to emphasize that it doesn’t actually think) etc those things are as they used to be and they don’t have this specific hallucination issue. The hallucination is really problem of the LLMs, but looks like now people call everything AI, which technically is not wrong (even things that we take for granted like GPS navigation or speech recognition are AI), but that help LLM obscure itself with other intentions.
I actually hope dentists use the old proven image recognition not asking chatgpt which is a chat bot that also incorporates other technologies to make it look better than it actually is.