?? Inside Wayve’s self-driving car as it tackles awful traffic
Exponential View via DALL-E

?? Inside Wayve’s self-driving car as it tackles awful traffic

A couple of weeks ago, I joined AI researcher turned co-founder and CEO of autonomous driving startup Wayve , Alex Kendall , for a test drive on a really difficult road in London. Wayve has been on my mind for a long time, and we wrote about it in EV#443 in the context of their most recent achievement — Alex’s team has created a 9-billion parameter world model called GAIA-1 to train their autonomous driving systems. Wayve uses synthetic training data, AI-generated driving situations, and prompting to enhance the driving capabilities.

I was blown away by the experience in the passenger seat. We drove on some of the toughest roads in London, and the system navigated the traffic — humans, cars, roadworks and even a police intervention — safely, smoothly, and with no human input.

See it for yourself in this brief video of our ride:

Alex and I later went back to his office where we recorded an hour-long conversation going under the hood of the technology, and Wayve's business model. You can listen to our full conversation here:

We cover:

  • ??? Alex’s journey from being a researcher to pioneering AI in transportation, inspired by groundbreaking AI achievements in other fields.
  • ??? The shift from traditional rule-based AI systems to dynamic, data-driven, end-to-end learning models in autonomous vehicles, mirroring trends seen in LLMs like GPT-3.
  • ??? “Embodied AI” and its role in shaping future everyday assistance, beyond just digital interactions.
  • ???Implementing AI in complex, unstructured environments like road traffic, and why Wayve’s approach using deep learning is paving a new path.
  • ??? The significance of scaling laws in AI development, and how they enable autonomous vehicles to handle edge cases and improve performance exponentially.
  • ???Alex’s perspective on Level 5 automation (fully autonomous vehicles) and why that might be a distraction to the opportunity.
  • ???The evolving business models in the automotive industry, shifting towards AI integration and subscription-based services.

This conversation isn’t just a technical deep dive – it’s a glimpse into the future of travel and urban living, shaped by AI.

LISTEN NOW

Brandon Hunter

Cloud Networking Professional

11 个月

The other day I saw a YouTube video of a police shootout on a highway in Houston, Texas. And countless police pursuits in Los Angeles, California among other places. I’d be curious to see how the AI’s handle those types of more chaotic driving scenarios where “the right thing to do” isn’t always very straight forward.

Javier Diaz-Gimenez

Professor of Economics at IESE Business School

11 个月

Amazing!

Muhammad Nawaz

Academic, Author of the NEH (Nawaz, Enscore, Ham) Algorithm

11 个月

The NEH (Nawaz, Enscore, Ham) algorithm is helping in self-driving vehicle research. please check this. Self-Driving Vehicle?Data Scheduling in Edge-Clouds B Burtchell, M Finch,?X Chen?- …?on Mobile Ad Hoc and Smart?…, 2022 - ieeexplore.ieee.org

回复
Peter Ac

Head of Innovations - Supply Chain at Inter IKEA Group

11 个月

Thank you for sharing. What kind of sensors are used to scan the physical world?

回复
Steven Spieczny

Kognic | Accelerating Embodied AI

11 个月

That looks like fun drive! d'ming you over email if you can check in.

回复

要查看或添加评论,请登录

社区洞察

其他会员也浏览了