@twoframesperminute lol 😂 they already listen including big brother. 😉
Conversation
Notices
-
nixCraft 🐧 (nixcraft@mastodon.social)'s status on Monday, 04-Dec-2023 15:33:51 JST nixCraft 🐧 -
nixCraft 🐧 (nixcraft@mastodon.social)'s status on Monday, 04-Dec-2023 15:28:36 JST nixCraft 🐧 Google and Apple can use AI to fight text and call scammers quickly. It can save money. Take advantage of advanced technology to help protect everyone instead of writing poems and creating videos using AI. It will have a real impact on securing communication. They can detect scammers in real time because technology has now improved so much. LLM is a perfect tool for the job. So please build these tools in iOS and Android.
-
Nazo (nazokiyoubinbou@mastodon.social)'s status on Monday, 04-Dec-2023 15:51:48 JST Nazo @nixCraft No, LLMs actually suck at identifying things. In most scenarios they're wrong more than 50% of the time. In a few they're wrong more than 90% of the time.
It's very important to remember that LLMs are not actually AI and lack the fundamental part of the AI -- the "intelligence" part. They don't understand, they just pattern. They'll improve filters and etc to try to force the LLMs in line, but ultimately it's just reinventing the wheel they already had with more energy waste.
-
Nazo (nazokiyoubinbou@mastodon.social)'s status on Monday, 04-Dec-2023 15:53:25 JST Nazo @nixCraft @twoframesperminute Maybe that's not ok? They actually don't though. You have quite a lot of control over things generally and a lot of stuff is opt in. I admit some is opt out, but opt out is not the same as no choice. What's more, if you have the knowhow (any Linux user should!) you can take even more control of your devices with third party "ROMs" when it comes to Android OS. (See LineageOS.)
We shouldn't encourage more listening in, we should demand less.
-