Awesome insight! Thanks for the brainstorming session. Also love that dude was doing the interview from inside his Model X. My 2022 MX already has 109,000+ miles. (I’m an Uber driver and love going on roadtrips)
Well, I personally think your conversations on Brainstorm is intriguing. This one is way over my head to worry about how AI software will evolve. But still interesting.
Good discussion trains; food for thought on what is defensible - a pay for performance where the performance is really dictated by who has the best/most/unique data (ie Tesla) starting with fremium to collect data and hook the customer; Then question becomes on the AI Sales agent, doesn't this eventually run into ethics constraints and gets regulated OR devolves to what we have today with robocalling and a counter technology just learns to block and filter it away...?
The AWS example is not a great one because converting from one java version to the next is really mainly about testing. The upgrade from 11 to 17 was really only minor which is opening up the module dependencies, which is very repetitive and could have been done by a developer writing a script based on the errors returned when running the app. It really wasn't an advanced use of AI in software development.
I was waiting for you to say that Tesla will be utilizing 'real-world' AI, not just selling software but building a tangible product that is "baby AGI" as Elon says. FSD embedded in a Tesla will be very hard for anyone to reproduce because of the data needed and the physical car built at scale.
![[Full] 4 ต่อ 4 Celebrity EP.922 | 10 พ.ย. 67 | one31](http://i.ytimg.com/vi/zmV5HRIxIBI/mqdefault.jpg)
Awesome insight! Thanks for the brainstorming session. Also love that dude was doing the interview from inside his Model X. My 2022 MX already has 109,000+ miles. (I’m an Uber driver and love going on roadtrips)
Well, I personally think your conversations on Brainstorm is intriguing. This one is way over my head to worry about how AI software will evolve. But still interesting.
Good discussion trains; food for thought on what is defensible - a pay for performance where the performance is really dictated by who has the best/most/unique data (ie Tesla) starting with fremium to collect data and hook the customer; Then question becomes on the AI Sales agent, doesn't this eventually run into ethics constraints and gets regulated OR devolves to what we have today with robocalling and a counter technology just learns to block and filter it away...?
The AWS example is not a great one because converting from one java version to the next is really mainly about testing. The upgrade from 11 to 17 was really only minor which is opening up the module dependencies, which is very repetitive and could have been done by a developer writing a script based on the errors returned when running the app. It really wasn't an advanced use of AI in software development.
I was waiting for you to say that Tesla will be utilizing 'real-world' AI, not just selling software but building a tangible product that is "baby AGI" as Elon says. FSD embedded in a Tesla will be very hard for anyone to reproduce because of the data needed and the physical car built at scale.
How these people have jobs?
are you guys on drugs while doing these videos?
Guys, guys, guys. "The explosion potential of software"? Come on. You can do better.
They’re ark look at their funds performance maybe they can’t lmao
wow. shallow!
First
nice