Transcript: YK Sugi on Reliable AI — Interview with Alexy

My name is YK. I'm a developer experience manager at Eventual. So far, being able to, you know, get to travel to different places like at Germany for the first time uh there speaking there about AI. When I think of a reliable system, it's a system that that's never down. I I would hope that for a reliable AI too I would you know I would want it to be never down and I would want like the whole ecosystem to work not just like the AI part of it but also the apps I don't want the apps to crash or make you know stupid mistakes just kind of basic stuff like that I think there there are a few things around it you know as I said an ecosystem around models themselves and and the models themselves too. But you know the mo like the models when it comes to that the open source ones are getting better. Um so basically it's a matter of like being able to serve it consistently being able to have a server you know like backup server and you know um being able to have a good kind of front end around it like whether it's a terminal UI or mobile app whatever it is make sure the code is good you know well tested basically you know good engineering practices it's it's always hard to predict the future I mean I feel like if if I watch myself in five years I might look stupid if I try to make a guess, but I would guess you know more more multimodal. I feel like that's that's fair to guess like more video, you know, input video output. Uh, I hope more reliable output too. You know, less hallucination and kind of more more grounded results.