Speaker unverified: evaluation

Transcript context

…And the second, mental health support made dramatically more accessible. So tell me more about the. architecture, the safeguards. I mean, we know the basics in terms of content filtering, classifiers to raise alarm bells when needed. But I guess my, I haven't really used chatbots much for any sort of therapeutic purpose, but my naive sense would be that, they probably do a pretty good job doing cognitive behavioral therapy, out-of-the-box and that, I think a lot of people are using them for that. Where do you find they fall short and tell me more about what you've built that isn't immediately visible to the user to improve on those weaknesses. Obviously, like a lot of companies would want to build something for their patient, for their customers using APIs like opening an API or I forward API, but they simply cannot because they do not have an in-house expertise to steer those model towards working the way it would want them to be. And then it's like this regulatory risk that you would have to address there. I hate risk of course. I mean, yeah, if you can finish feeling kind of right backed about that, his country spans around 70% of corporates on safety. It's right. It would be something like very, very expensive. I generally be on the technical level where we have a bunch of classifiers that are constantly running on the background. So it's in real time, but ultimately right now, especially when he needs to feel longer about certainly stood for about two days. I'm having that we are quiet that it it's a humor to cloud conflict because it exactly about you in different tools and. For example, if you want to plan to benefit on my head. Or do they're blacked on what happened in the bad trick, right? Like I mean from the channel perspective or from the agent perspective, what happened in the last common day and the conversation with this user and how this would inform or plan and our actions going forward. So the agent is able to deploy those or maybe sub agents that do talk like that and then and the feed guided information into the main context and then it continues. So under a lot of that you have this like play powerful memory and the very powerful planning capabilities that compare. It also has the chances that it will meet. Sometimes they even call them for the user safety. They are dramatically lower than that. But then with all this attitudes. How are people accessing this? I guess maybe one more just to wrap us up. Are there any anecdotes from your deployments in Ukraine or in U.S. prison populations that you think are particularly memorable or inspiring that you would leave people with, or failing that, just anything else you would want to leave people with in terms of a positive vision of the future.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.