Tristan Rice (@rice_fry), a software engineer specializing in Machine Learning, Reverse Engineering, and Security, shared a pretty interesting render of the 3D voxel models used by Tesla’s Full Self-Driving (FSD) beta 10.5 on Twitter Wednesday evening.
It seems to have no issues with close pillars/walls even in tight parking spaces or directly facing into a wall.
— Tristan (@rice_fry) November 25, 2021
Tesla started rolling out FSD beta 10.5 earlier this week with the 2021.36.8.8 software update.
The render demonstrates how robust the 3D modelling techniques used by FSD beta 10.5 are, and how they are able to accurately map tight spaces like a congested parking garage even though FSD doesn’t actually work in indoor settings like parking garages.
Tesla CEO Elon Musk responded to Tristan’s tweet, offering the clever software engineer a job working on Full Self-Driving. “Why watch it happen instead of helping making it happen?” wrote Musk on Twitter.
Why watch it happen instead of helping making it happen?
— Elon Musk (@elonmusk) November 25, 2021
You know you’ve got something of value when the CEO of a trillion-dollar company personally offers you a job over Twitter.
Tesla is always on the lookout for new talent in the Artificial Intelligence and Machine Learning spaces to help flesh out the electric vehicle (EV) maker’s vision for autonomous driving software.
Unfortunately, Tristan didn’t seem too keen on taking Musk up on his offer even though he found it tempting. “The offer is very tempting–I really like all the folks I’ve met on Autopilot and it’s a very interesting problem,” wrote the software engineer.
That's definitely part of it. The offer is very tempting–I really like all the folks I've met on Autopilot and it's a very interesting problem.
I've been having a lot of fun working in open source with PyTorch (indirectly Tesla) and posting stuff like this as a hobby
— Tristan (@rice_fry) November 26, 2021
Rice currently is a Senior Software Engineer at Meta, the parent company of Facebook, working out of Seattle, Washington.
“I work on making large scale machine learning more reliable, efficient and scalable. Distributed model and data parallel training. The largest Instagram, Facebook, and Ads models are trained on the software I work on,” details Rice’s LinkedIn page.
Tristan ceded that he’s just been “having a lot of fun working in open source with PyTorch.” Since Tesla uses PyTorch for a number of purposes, that already puts the software engineer in close proximity with Tesla’s work.
He also wouldn’t get to post cool renders like the ones in his original tweet if he hopped over to Tesla, which he said was part of why he didn’t want to make the move at this time.
Uncovering Tesla tidbits isn’t knew to the Meta senior software engineer. Back in April, Rice uncovered what Tesla Insurances knows while you’re driving, as he shared on Twitter a list of telemetry data used by the electric automaker.
Nehal has a passion for everything tech — from electric vehicles to the latest smartphones, and everything in between. Click here to get 1,000 Tesla Supercharger miles free with your next Tesla purchase.