The surprisingly boring road to self-driving cars

Comment

waymo hybrid chrysler pacifica
Image Credits: Dllu (opens in a new window) / Wikimedia Commons (opens in a new window) under a CC BY-SA 4.0 (opens in a new window) license.

At last, it is here! The truly self-driving car, no human behind the wheel! For the public! …A few hundred of them, in a closed beta, in a small corner of sun-drenched (never snow-drenched, almost never water-drenched) suburban Phoenix, five years later than some people were predicting six years ago.

Few new technologies have ever been more anticipated and more predicted than the self-driving car. Anyone who drives cannot help but imagine not having to drive any more. It has been said that they will change our cities, our homes, our commerce, even our fundamental way of life.

But at the same time, the actual progress has seemed … well … glacial, to the casual driver’s eye. We’re mostly talking about software, after all. OK, and LIDAR, and cameras, but the software is the key. People couldn’t help but expect a roll-out like that of smartphones, where the launch of the iPhone in 2007 led to adoption by every tech-savvy person by 2010, and the vast majority of the developed world by 2013.

People couldn’t help but expect a mass market push. In 2014, the optimistic attitude was, maybe your next car is electric; then your next one — or even that same one, courtesy of an OTA software update — will be self-driving! Set the controls for the heart of Los Angeles, or Boston, or both, and lie back and snooze, baby.

That’s not how it’s going to happen. Waymo’s closed beta is a huge step, yes, but it is also a tiny incremental iteration. We aren’t going to see a Big Bang moment, when suddenly you buy your next car and it will carry you unaided from Vancouver to Halifax, or even Vancouver to Whistler. Instead we’re going to see a series of tiny steps forward, measured over years, frequently in industrial or commercial settings rather than personal ones.

First they drive the broad, sunny streets of Phoenix; then highways; then in more complex situations, such as airports and downtowns; then in heavy rain; then amid detours and road closures; then in rough, winding country roads prone to landslides and flooding; then (some considerable time from now, says your Canadian correspondent) in snow and ice…

And even then, how can a truly self-driving car handle anomalous situations, when the car doesn’t know what to do and screeches to a halt? Even more importantly, how will it know it’s in an anomalous situation and it doesn’t know what to do? Will cars be drivable remotely, in such cases? If so, how will we secure that process? What about adversarial attempts to manipulate the neural networks behind the figurative wheel, by feeding them misleading inputs that they respond to but the naked human eye might not notice?

I suppose we have to talk about the so-called “trolley problem,” too. I’d rather not. It is by far the silliest and most overanalyzed question about self-driving, since in 99.9% of problematic situations the solution is simply “stop.” Anything like the trolley problem will only come up in the edgiest of edge cases — but, if only to satisfy the public, those cases will have to be publicly hashed out as well.

The larger issue brought up by the “trolley problem problem” is that we have no collective social understanding of how to judge the risks posed by self-driving cars, and what risks we should accept. On paper, if all of America moved to self-driving cars overnight and they started killing 100 people every single day … America should rejoice, because the death rate from car crashes will have fallen!

In practice, however, he understated, it seems likely that America, or at least American media, will not rejoice. Rather the opposite.

When you step into a self-driving vehicle, you will be taking a risk, just as you do whenever you step into a human-driven vehicle. But it will be harder to measure this new risk, and even if/when we can, we won’t weigh it the same way that we do the old risk. Such is human nature. Liability alone will be a giant can of worms.

We have an entire infrastructure of regulation built around the old risk. It will change only slowly to manage this new risk, and it will have great difficulty sloughing off old preconceptions which no longer apply. Dream of cars with no steering wheels all you like, for example, but my guess is that in many jurisdictions, self-driving cars will have to include a legal driver among their passengers at all times.

When you consider the combination of the technological challenges, the social challenges, and the regulatory challenges, all of which are seriously nontrivial — it seems apparent that we are going to creep, rather than bound, into the self-driving future.

And so: self-driving vehicles will slowly, quietly, take over closed industrial / commercial settings. Waymo’s self-driving taxis, followed (apparently at some distance) by others, will very gradually expand their beachhead from Phoenix, bit by bit and clime by clime, with occasional setbacks. Personal cars will continue to increase their self-driving capabilities one situation at a time: parallel parking, stop-and-go highway traffic, parking garages, certain patches of quiet suburban territory.

This means there will almost certainly be no point at which you suddenly have a self-driving car. Self-driving isn’t a product, an event, or a feature; it’s an aspirational limit to which we will asymptotically approach. We’re collectively already on that curve — which is exciting! — but it seems apparent that its climb will be much more gradual than almost everyone, including me, thought not so long ago.

More TechCrunch

For over six decades, the nonprofit has been active in the financial services sector.

Accion’s new $152.5M fund will back financial institutions serving small businesses globally

Meta’s newest social network, Threads is starting its own fact-checking program after piggybacking on Instagram and Facebook’s network for a few months. Instagram head Adam Mosseri noted that the company…

Threads finally starts its own fact-checking program

Looking Glass makes trippy-looking mixed-reality screens that make things look 3D without the need of special glasses. Today, it launches a pair of new displays, including a 16-inch mode that…

Looking Glass launches new 3D displays

Replacing Sutskever is Jakub Pachocki, OpenAI’s director of research.

Ilya Sutskever, OpenAI co-founder and longtime chief scientist, departs

Intuitive Machines made history when it became the first private company to land a spacecraft on the moon, so it makes sense to adapt that tech for Mars.

Intuitive Machines wants to help NASA return samples from Mars

As Google revamps itself for the AI era, offering AI overviews within its search results, the company is introducing a new way to filter for just text-based links. With the…

Google adds ‘Web’ search filter for showing old-school text links as AI rolls out

Blue Origin’s New Shepard rocket will take a crew to suborbital space for the first time in nearly two years later this month, the company announced on Tuesday.  The NS-25…

Blue Origin to resume crewed New Shepard launches on May 19

This will enable developers to use the on-device model to power their own AI features.

Google is building its Gemini Nano AI model into Chrome on the desktop

It ran 110 minutes, but Google managed to reference AI a whopping 121 times during Google I/O 2024 (by its own count). CEO Sundar Pichai referenced the figure to wrap…

Google mentioned ‘AI’ 120+ times during its I/O keynote

Firebase Genkit is an open source framework that enables developers to quickly build AI into new and existing applications.

Google launches Firebase Genkit, a new open source framework for building AI-powered apps

In the coming months, Google says it will open up the Gemini Nano model to more developers.

Patreon and Grammarly are already experimenting with Gemini Nano, says Google

As part of the update, Reddit also launched a dedicated AMA tab within the web post composer.

Reddit introduces new tools for ‘Ask Me Anything,’ its Q&A feature

Here are quick hits of the biggest news from the keynote as they are announced.

Google I/O 2024: Here’s everything Google just announced

LearnLM is already powering features across Google products, including in YouTube, Google’s Gemini apps, Google Search and Google Classroom.

LearnLM is Google’s new family of AI models for education

The official launch comes almost a year after YouTube began experimenting with AI-generated quizzes on its mobile app. 

Google is bringing AI-generated quizzes to academic videos on YouTube

Around 550 employees across autonomous vehicle company Motional have been laid off, according to information taken from WARN notice filings and sources at the company.  Earlier this week, TechCrunch reported…

Motional cut about 550 employees, around 40%, in recent restructuring, sources say

The keynote kicks off at 10 a.m. PT on Tuesday and will offer glimpses into the latest versions of Android, Wear OS and Android TV.

Google I/O 2024: Watch all of the AI, Android reveals

Google Play has a new discovery feature for apps, new ways to acquire users, updates to Play Points, and other enhancements to developer-facing tools.

Google Play preps a new full-screen app discovery feature and adds more developer tools

Soon, Android users will be able to drag and drop AI-generated images directly into their Gmail, Google Messages and other apps.

Gemini on Android becomes more capable and works with Gmail, Messages, YouTube and more

Veo can capture different visual and cinematic styles, including shots of landscapes and timelapses, and make edits and adjustments to already-generated footage.

Google Veo, a serious swing at AI-generated video, debuts at Google I/O 2024

In addition to the body of the emails themselves, the feature will also be able to analyze attachments, like PDFs.

Gemini comes to Gmail to summarize, draft emails, and more

The summaries are created based on Gemini’s analysis of insights from Google Maps’ community of more than 300 million contributors.

Google is bringing Gemini capabilities to Google Maps Platform

Google says that over 100,000 developers already tried the service.

Project IDX, Google’s next-gen IDE, is now in open beta

The system effectively listens for “conversation patterns commonly associated with scams” in-real time. 

Google will use Gemini to detect scams during calls

The standard Gemma models were only available in 2 billion and 7 billion parameter versions, making this quite a step up.

Google announces Gemma 2, a 27B-parameter version of its open model, launching in June

This is a great example of a company using generative AI to open its software to more users.

Google TalkBack will use Gemini to describe images for blind people

Google’s Circle to Search feature will now be able to solve more complex problems across psychics and math word problems. 

Circle to Search is now a better homework helper

People can now search using a video they upload combined with a text query to get an AI overview of the answers they need.

Google experiments with using video to search, thanks to Gemini AI

A search results page based on generative AI as its ranking mechanism will have wide-reaching consequences for online publishers.

Google will soon start using GenAI to organize some search results pages

Google has built a custom Gemini model for search to combine real-time information, Google’s ranking, long context and multimodal features.

Google is adding more AI to its search results