AV

This car is practicing how to say “dalé” and “y’all”

Argo AI is bringing a bunch of these to the Lyft app in Miami this year and Austin shortly after. The self-driving startup plans to deploy at least 1,000 Fords on the Lyft network over the next five years.

Rather than reinvent the wheel (and car) from scratch...

...Argo partnered with automakers from the get-go, integrating its stack with cars made by Ford and Volkswagen, its two big investors. Feast your eyes on this Argo/VW collab: 

Argo

 “We always knew that we would be partnering on both the supply of vehicles and the demand part of the equation,” Argo CEO Bryan Salesky told the Brew. Argo struck quite a unique deal with its new demand partner. 

Lyft is getting a 2.5% equity stake in Argo (worth ~$310 million, per Bloomberg). In return, the startup gets “two large corpuses of data,” per Salesky:

1. Aggregate safety performance of Lyft drivers on US roads. On a granular urban level, this could shape where Argo’s vehicles do and don’t go.  
2. Demand patterns and “trip movements” data, highlighting “real transportation issues where there may not be enough supply of Lyft drivers on the network.” Argo could backfill with robotaxis in these places and use the data to determine new target markets. 

On commercialization, Argo is cargo-agnostic: It’s open to people and packages. The company has pilots running in the “goods movement space,” but won’t announce them yet. 

Speaking from experience

Miami and Austin aren’t terra incognita for Argo. Its AV test fleet is already pounding the pavement in both the Magic City and the capital of Texas. 

“They’re both challenging cities to operate in,” Salesky said. “You get very pedestrian-intense areas where nobody’s following the rules [on] crosswalks and roads in Miami Beach. ...And then in Austin, you have this huge scooter population.” This ATX resident can confirm that last observation.—RD

SEMICONDUCTORS

How to lighten up a datacenter

What if we took the tech powering quantum computing and retrofitted it to AI processing? 

Many years ago, Nick Harris, cofounder and CEO of Lightmatter, asked himself that question. He was a doctoral student at MIT, working on quantum computers, and realized he could repurpose the way they process light to solve a problem with another technology he was deeply familiar with: semiconductors. 

So in 2017, Harris founded Lightmatter to marry his quantum experience with his semiconductor knowledge and apply photonic computing, a method of using light in AI processing, to make chips faster and more energy efficient. 

• Lightmatter’s processors are built for data centers, and it says its AI chips are 10x faster than Nvidia and other leading chipmakers’ similar offerings. 
• In May, Lightmatter announced an $80 million funding round from investors including HP Enterprise and Lockheed Martin.

How it works: “On an electronic chip, when you want to send a signal between points, you have to turn a wire off and on. And turning a wire off and on costs energy,” Harris told us. “With an optical wire, you don't have those properties. So turning it off and on takes very, very little energy...and that allows you to run the processor at very high clock speeds.”

We chatted with Harris about the startup’s biggest challenge, its go-to-market strategy, and how its chips could affect everything from Siri responses to your next retail shopping experience. 

Click here to read the full interview.—HF

SPONSORED BY CROWDSTRIKE

Frustrated by Antivirus Scans That Move Like Molasses?

What about endless updates, reboots, alert fatigue, and missed threats? If you’re into that kinda stuff, no need to keep reading. 

As for the rest of us (meaning all of us), wouldn’t it be great to say “arrivederci” to these productivity killers? 

Well now you can. Say farewell to your ancient antivirus software and “bonjourno” to your new cybersecurity platform, CrowdStrike.

CrowdStrike offers super fast, super easy-to-use antivirus scans, endpoint protection, and more.

It won’t slow your machines down and never needs to be updated. It’s just fast, easy, and super secure—which means no more complaints about reboots and lag time—just happy employees. 

The only thing better than saying “byeeee” to your ancient cybersecurity software is trying Crowdstrike for free.

Start your free trial today. 

SEMICONDUCTORS

Supersized subsidies

In the recent semiconductor-shortage-dominated months, the US, European Union, and South Korean governments have pledged to boost their chipmaking subsidies from basically (or, in the US’ case, literally) zero to hundreds of billions of dollars.  

• For context...here’s each country’s share of global semiconductor manufacturing: US (12%), EU (2.8%), South Korea (27%). 

“We’ve sort of reached a tipping point where countries are investing because they know demand for semiconductors is on the rise,” Dan Rosso, a Semiconductor Industry Association (SIA) spokesperson told us. “They want a bigger share of the increased chip production.”

Big spenders

So far in 2021, the US, EU, and South Korean governments combined have pledged 5,124% more in semiconductor subsidies than they dished out over the last two decades.

• From 2000–2020, total semiconductor subsidies across South Korea, the US, and Europe totaled ~$12.5 billion, per SIA analysis, with the US spending zero, zip, zilch, nada. 
• Now, that figure is ~$653 billion. South Korea is the leader with a plan to spend $451 billion, the EU is next at a pledged $150 billion, and the US Senate approved a plan to spend $52 billion. 

Zoom out: China, which has 12% of global semiconductor manufacturing, pumped about $50 billion into its chipmaking industry between 2000–2020. And in 2014, it established a goal of spending $150 billion to produce 70+% of the chips it uses by 2030. As of H1 2021, it made no more than 36% of the chips it used.

BITS & BYTES

Stat: Compared to last year, US electric vehicle sales more than doubled in H1 2021—though they’re still just a small chunk of the overall auto market. Tesla drove the majority of that growth.  

Quote: “I think a good vision for the metaverse is not one that a specific company builds, but it has to have the sense of interoperability and portability.”—Mark Zuckerberg to The Verge in a new, metaverse-focused interview

Read: Can trusty ol’ radar—not lidar—help autonomous vehicles see? 

Fantasy stock market. When you play MarketWatch’s Virtual Stock Exchange, you build a portfolio and compete against your friends or coworkers while you’re at it. Earn bragging rights when you get started with MarketWatch today.*

SPONSORED BY WEBEX BY CISCO

When it comes to speed, trust the experts. The McLaren F1 team uses Webex for virtual, real-time communication that can keep up with their pace. Learn more about how Webex can ready your team for this new era of hybrid work here.

WHAT ELSE IS BREWING

• SpaceX owns bitcoin, Elon said Wednesday. Elon himself owns bitcoin, ether, and dogecoin.  
• Ericsson is worried it will be locked out of the Chinese market following a Swedish ban of Huawei 5G networking gear. 
• Waymo is opening an office on Argo’s turf (Pittsburgh). 
• Aramco was hit by ransomware. 
• Form Energy, a four-year-old battery startup, claims it has created an inexpensive, high-capacity battery based on an abundant material: iron. 
• Clearview AI—the controversial facial recognition company—has raised $30 million in a new round of funding, from investors who, ironically, wish to remain unidentified. 

GOING PHISHING

Three of the following news stories are true, and one...we made up. Can you spot the odd one out?

1. White Claw's parent company made a robotic cooler that follows you around.
2. A surprise guest joined Ashton Kutcher and Mila Kunis to discuss the ABCs of crypto. 
3. The UAE is using drones to control its weather.
4. Twitter is testing a dislike button with a limited amount of iOS users. 

MARKET RESEARCH

TikTok is famously secretive about its too-accurate, wormhole-inducing algorithm. Alas! A new, inventive, bot-based investigation from the Wall Street Journal gives us a glimpse under the hood. 

WSJ assigned 100 bots to “interests” (e.g., sports or animals). They didn’t disclose this information to TikTok, but rather programmed the bots to watch videos that were hashtagged to their interests. Through this experiment, the WSJ learned that…

1. Watch time/pause time is TikTok’s key metric.
2. It takes the app very little time to get to “know” you. 
3. TikTok, like other social media platforms, will feed you more and more of the same content if you express an interest.

Check out the video here.

ICYMI

Catch up on the top Emerging Tech Brew stories from the past few editions: 

• Oculii, the autonomous vehicle vision company, has a big new partnership
• Call centers eye AI for automation, emotion analysis
• What it’s like to take a fully driverless robotaxi

GOING PHISHING ANSWER

White Claw has not come out with an answer to Heineken's autonomous cooler.