Recently saw Zoox published their first an hour fully autonomous driving video, the result is really impressive! Cruise also published this kinds of video before, but it seems like it does not draw lots of attention. Both company focuses on city driving in SF. Which company do you think has better technology in terms of fully autonomous driving? Cruise an hour autonomous driving: https://www.youtube.com/watch?v=LCNLOJS39V8 Zoox an hour autonomous driving: https://www.youtube.com/watch?v=6r7vDhPXmiM #auto #career #cruise #zoox
What's the background one needs to have to work in the AV industry? Computer Vision, Deep Learning, Perception? What's the best way to educate oneself? Masters? Online course?
None
Zoox has a marketing budget that exceeds their engineering budget. Cruise has a lot more funding and industry support with _a lot_ more vehicles on the road. Vehicles on the road = data = find more edge cases = find solutions for the 10% of issues that hold up large scale AV deployment.
I'm not sure that's true lol. I mean, the last part is facts. But I'm not sure we have much of a marketing budget lol
Yeah the comment about marketing is really stupid and uneducated. We do almost zero marketing, except for videos on Twitter and YouTube. You think those videos cost anything compared to staffing 900+ employees? 🤦🏼♂️
I had to look into this a lot when choosing between offers and overall I came to the conclusion, which I still think is true today from the inside, that Zoox has the better driving AI but worse investment in other crucial assets like infra and fleet scaling. Cruise is the opposite. Cruise has a lot of built up infra, fleet scaling, learnings from high miles and dogfooding their product, but overall even their sizzle real videos have a lot of issues. Planner worms all over the road. Lots of harsh braking (you can observe this if you walk too closely to a crosswalk when one of the Bolts is driving around). They seem to drive in last interesting areas, though I'm seeing cruise more and more in Zoox's turf so to speak. It's hard to compare ODDs but I think Zoox has an edge in driving for now, but really depends on if Cruise's investments in infra pay & miles pay huge dividends over the next two years and if Zoox can manage the same or not.
Cruise AV visualizations look awful compared to Zoox. I’m giving this one to Zoox. Ride commentary and their overall production of videos seem more interesting.
I don't think either company has published their full internal visualizations. The zoox commentary was nice. Cruise has open sourced a lot of our viz tooling although I don't think we have published any detailed videos showing how powerful it is.
Neither
Argo.
Not Cruise
You are comparing Cruise driving on San Francisco street since 2016 vs Zoox driving in 2020 when there are no other cars on the street due to quarantine
Did you watch the video, it was mentioned at the begining that it happend several months ago and you can see a lot traffic participants.
Yeah, that drive happened months ago. I don't even know if Zoox has been testing at all since the shelter-in-place went into effect.
The problem with autonomous driving is not the 90% of the time when it works, but the 10% of the time when it doesn’t
At this point it's more like the .1% that it doesn't, and reducing every day
Have you ever considered maybe we're solving the problem the wrong way. At this point it would of been cheaper to lay high speed rails.