Faster, Cheaper, Better AI

Happy Thursday.

The Dow Jones Industrial Average has passed the 28,000 mark, apparently prompted by news that China and the U.S. are zeroing in on a trade deal.

And to help with translation of any future meetings between the parties, they might want to consider the Google real-time translation that works in conjunction with Google Assistant now on mobile devices.

The new feature is now available in 44 languages, and should help you more easily ask directions to the nearest metro station around the globe.

The AI juggernaut of apparently continued unabated this year, according to the AI Index Report, which reported a 300% rise in research papers.

The report generally aspires to get a handle on AI progress, yet another headline seems to be that while AI is on the march, we’re a long way from reaching “general intelligence.”

The U.S. continues to lead in terms of private investment (just under $12B, compared to China’s $6.8B, which places them in second).

Algos are coming faster, cheaper, smarter, easier to train, and self-driving cars continue to get more investment than in any other field.

I eagerly await the algo and/or SDV that can cleverly route me from north to south Austin in rush hour without raising my blood pressure!

New (Google Ad) Rules

Google is jumping on the political advertising guard rail bandwagon, announcing new limits on its microtargeting to age, gender, and general location (zip code). Google will also continue to allow retargeting based on content of website visits.

They will no longer allow ads to be directed to specific audiences based on public voting records or political affiliations (“left-leaning,” “right-leaning,” etc.)

The policies will impact both Google search results, on YouTube, and via the Google Content Network which displays ads on other sites.

The move left many politicos stunned, for in past campaigns they have leaned mightily on microtargeting specific sets of voters, an efficiency that makes many grassroots campaigns financially feasible.

Also on the GOOG front…remember Google Duplex, the creepy-crawly AI that would sound like a human and call to make your hair appointments? Well, Duplex is now officially moving beyond the confines of voice ops and launching as “Google Assistant in Chrome” as a streamlined workflow to help you buy movie tickets.

Next up: Streamlining the process of renting a car. And then?…maybe using Google Duplex to more easily buy political search ads from Google??!

Making Copies

Happy Monday.

Only 10 days to Turkey Day. Get ready to gobble gobble.

I’ll tell you who’s not making copies. That would be HP, the board of which rejected Xerox’s acquisition bid over the weekend claiming the $22/share bid undervalued HP and was not in the best interest of shareholders.

Operators of Yahoo sites in Japan and the Line chat app have reached agreement on a merger, one where SoftBank Group and SK internet firm Naver will form a joint holding company. The new entity is expected to become the largest internet player in Japan.

On the funding front: AI-driven news app SmartNews has raised a $92M round at a valuation of $1.2B. The app is said to use machine learning (rather than human curation) to generate personalized news experiences.

The company is focused initially on Japan and the U.S. as key markets.

Self-driving startup Wayve has raised $20M in a Series A to help teach cars to drive autonomously using reinforcement learning, simulation, and computer vision.

Wayve is looking to improve AI through simulation and knowledge transfer to the real world, rather than “hand-coded” rules. It plans to launch a pilot fleet of autonomous Jaguar I-Pace cars in Central London which will include, for now, safety drivers. 

Bad Weather

Happy Friday, and TGIF.

Yesterday, some news about the weather from IBM with the introduction of its GRAF weather forecasting model.

GRAF stands for “Global High-Resolution Atmospheric Forecasting” and is the first global weather model to run on a GPU-accelerated supercomputer called “DYEUS.”

In partnership with the National Center for Atmospheric Research, this new system uses state-of-the-art science to forecast the atmosphere on a global scale, providing hourly (vs every 6-12 hours) updates and bringing forecast fidelity once limited to a few countries to the rest of the world.

Forecasts from the new system will be available via The Weather Channel and Weather Underground apps and websites, and also to businesses via IBM offerings.

IBM GRAF runs on supercomputer powered by the IBM Power9 System and CPUs and GPUs optimized on open source software for AI and high-performance computing.

Also on the cloud front…Amazon is filing a protest of the Pentagon’s decision to award the JEDI contract to Microsoft, claiming “unmistakable bias” and “errors” in process.

And Apple is just saying no to vaping, announcing it will remove all 181 vaping-related apps from its store amid growing health concerns over e-cigarettes.

Streaming to the Max

Details are emerging on HBO Max, HBO’s new live streaming service.

It’s $15/month (U.S.), which is what existing HBO costs…but AT&T is planning to bundle free subscriptions for some customers of its other services.

Compare that to Apple’s $5/month and Disney’s $7/month (Netflix comes in between $9 and $16/month, depending on what flavor you get).

On the subject of money, Sony announced a Q2 operating profit of $2.56B, jumping 16% on its 102.8M total unit sales of the PlayStation 4 (now larger than the original PlayStation).

Funding Rounds: Duality, which makes privacy-preserving data analysis tools w/ homomorphic encryption, raised $16M in a Series A led by Intel Capital. And Quill raised a $2M seed and $12.5M Series A for its messaging product and Slack competitor.

At today’s TensorFlow World conference in Santa Cruz, Google launched TensorFlow Enterprise, an “optimized” version of its open source machine learning framework for large businesses.

Tonight: It’s the Houston Astros vs. the Washington Nationals in a winner-take-all game 7. On the mound, Zack Greinke for Houston & Max Scherzer for Washington. It’s hard to believe it came down to a game 7, but that’s late October for ya!

Play ball!

Big Builds, Bigger Dogs

Yesterday it was AI and agriculture, today it’s AI and construction. Built Robotics is a company looking to make construction equipment autonomous and has raised a $33M Series B round. Recognizing that the industry is facing a labor shortage, Built is building systems that would allow one equipment operator to oversee a fleet of vehicles working autonomously in parallel. 

But there’s plenty more VC money sloshing around where that came from…Wordpress parent Automattic raised $300M in a Series D round from Salesforce Ventures, valuing the company at $3B. Thirty-four percent of the world’s top 10 million websites now run on WordPress. No word yet if that means everyone at the company gets an “automattic” raise! 

There’s also more consolidation on the developer tooling front with GitHub’s acquisition of code analysis tool Semmle. Semmle streamlines security testing and offers developers a query languages to allow researchers to more easily test their code. No price on the deal, but Semmle was born from Oxford just last year and had raised a $21M Series B round led by Accel.

And app performance monitoring firm Datadog raised $648M in its U.S. IPO, valuing the company at $7.83B. Datadog had previously declined a buyout offer from Cisco.

FarmBots

If you thought automation was limited to e-commerce warehouses and factory floors, think again. TechCrunch is reporting that a number of robotics and autonomy firms are looking next to agriculture.

From farm to C3P0 to table.

According to a report from TechCrunch, Ag robotics firm FarmWise just raised $14.5M to continue development of an autonomous weeding vehicle.

Though they are currently prototypes, the large machines scan the ground for invasive weeds among crops and carefully pluck them out. 

From the company’s press release on the fund raise: “FarmWise’s autonomous, AI-enabled robots are designed to solve farmers’ most pressing challenges by performing a variety of farming functions — starting with weeding, and providing personalized care to every plant they touch.”

The company suggests that machine learning models, computer vision and high-precision mechanical tools will allow sophisticated robots to cleanly pick weeds from fields, leaving crops with the best opportunity to thrive while eliminating harmful chemical inputs.

FarmWise is supported by a team of 25 farming and AI experts from MIT, Stanford, and Columbia. 

Shelf Life

On Tuesday California legislators passed AB5, a landmark bill that requires companies like Uber and Lyft to treat contract workers as employees. Workers must be designated as employees if a company exerts control over how they perform their tasks or if their work is part of a company’s regular business.

Expect Uber and Lyft’s autonomous vehicle efforts to speed up (as well as their lobbying efforts to gain an exception for its drivers to remain contractors). All those pesky humans, demanding rights like fair wages and health insurance!

On the subject of robots, Simbe Robotics has raised $26 million for autonomous inventory robots to inventory grocery shelves. A VentureBeat story indicates the brick-and-mortar automation market will be worth $18.9B by 2023.

Simbe’s robot, Tally, drives around a space to create a store map and then uses computer vision to “see” what products aren’t on a shelf and identify any missing facings, using RFID for its inventory counts.

VentureBeat reports a single robot can scan 15K to 30K products per hour, compared to the 10K-20K an average human employee can do in 20 to 30 hours. Human, 10-20K, 20-30 hours, Tally, 15K-30K per hour.

Humans will still do the restocking…for now. 

We’ll see how long their shelf life is.

McAutomation

Ronald McDonald retired around the time of the creepy clown sightings began popping up several years ago. But McDonald’s isn’t resting on its fast food franchise laurels.

Yesterday the company announced it had agreed to acquire voice AI firm Apprente to help the company automate drive-thru orders (which could also cut down on service times).

According to a report from CNBC, Apprente’s employees will be founding members of “McD Tech Labs,” a tech shop embedded in Mickey D’s global tech team.

This move builds on the company’s investment in March of Dynamic Yield, whose ML technology was deployed in more than 8,000 U.S. drive-thrus, and whose tech changes menus based on time of day, weather, traffic, and the customer’s order, with the end goal of upselling.

Today’s Morning Brew suggested this was a “defensive” move by McDonald’s, “ensuring it gets access to key AI tech before Wendy’s and Burger King can get their hands on it.”

There’s a lot at stake in the coming AI burger wars…turnover is inordinately high at fast food chains, and McD’s alone employs over 1.8M workers in 34,000 restaurants worldwide. Those algos don’t complain about low wages, few benefits, and long hours.

So in my best speech-to-text voice, here goes: Would you like fries with that?

Closing the Widening Skills Gap

IBM’s Institute for Business Value released a study on Friday focused on the impact of AI on the workplace.

The study revealed over the next three years, as many as 120 million workers from the world’s largest economies may need to be retrained because of AI advances and intelligent automation.

But less than half of CEOs surveyed said they had the resources needed to close the skills gap brought on by these new technologies. And the time it takes to close a skills gap through training has increased by more than 10X in just four years.

The study also revealed that new skills requirements are quickly emerging, while others are becoming obsolete. In 2018, the two top skills were behavioral: A willingness to be flexible, agile, and adaptable to change, and time management skills and ability to prioritize.

How to close the gap? The core recommendation is to take a more holistic approach by focusing on reskilling our workforce through development that’s multi-modal, personalized to the individual and built on data, learning journeys that are delivered through “experiential learning.”

You can read a summary of the research here.