Google's Med-PaLM 2, one tool τεχνητής νοημοσύνης που έχει σχεδιαστεί για να απαντά σε ερωτήσεις σχετικά με ιατρικές πληροφορίες και βρίσκεται σε δοκιμή στο ερευνητικό hospital Mayo Clinic.
According to a Wall Street Journal article Med-PaLM 2, a variant of PaLM 2, is the language model on which Google's Bard is based, trained in medical science and tested since April. At Google they believe it can be particularly useful in countries with "more limited access to doctors".
Med-PaLM 2 was trained on a curated set of medical knowledge that Google believes will do better in healthcare conversations than generic chatbots like Bard, Bing and ChatGPT.
In research that Google published in May shows that Med-PaLM 2 still suffers from some of the accuracy issues we are already used to seeing in large language models.
At study, οι γιατροί βρήκαν ανακρίβειες και άσχετες πληροφορίες στις απαντήσεις που παρείχαν τα Med-PaLM και Med-PalM 2 της Google, σε σχέση με εκείνες των πραγματικών γιατρών.
According to Google's senior director of research, Greg Corrado, Med-PaLM 2 is still in its early stages, but is being tested in real-world conditions.
They are out of their minds