A Blog by Jonathan Low


Sep 21, 2017

Microsoft Unveils New System For Superfast Artificial Intelligence

AI is not just getting smarter, it's getting faster, which means real time insights - and Microsoft's emergence as a stronger force in artificial intelligence. JL

Blair Frank reports in Venture Beat:

Microsoft showed off a new system that will allow developers to deploy machine learning models onto programmable silicon and achieve high performance beyond what they’d be able to get from a CPU or GPU. FPGAs let programmers configure hardware optimized to execute functions prior to runtime, like doing the math to serve insights from neural networks. The lack of batching means it’s possible for the hardware to handle requests as they come in.“We call it real-time AI because the idea here is that you send in a request, you want the answer back.”

Can Ian Schrager's New NYC Hotel Beat Airbnb?

Selling the experience, not just the room. JL

Chris Kirkham reports in the Wall Street Journal:

“Airbnb is a mortal threat to the U.S. hotel industry,” said Mr. Schrager, known for creating Studio 54 in the 1970s and later the first wave of boutique hotels including the Royalton and Paramount. He aims to better compete with Airbnb on nightly rates and offer superior amenities such as bars and places to socialize, reaping 35% to 45% of  revenue from food and beverage, compared with a national average of 24%. He is aiming for profit margins of 45% at the bars and 25% at the restaurant—about 20% higher than typical at hotels.

Uber Sues Its Ad Agency For Fraudulent Advertising

The good news for Uber is that data underscore how strong its underlying, organic growth is. JL

Alison Griswold reports in Quartz:

Uber is suing for about $40 million for fraud, alleging it was actively misled as Fetch “squandered tens of millions of dollars to purchase nonexistent, nonviewable, and/or fraudulent advertising.” This is a relatively novel position for Uber, which is a named defendant in about 250 federal cases but rarely on the offensive. Uber says the number attributable to paid mobile ads fell, but the number of organic installations rose by almost the same amount. The entire complaint is basically an ad for how terrific Uber’s organic user growth is.

Designing Workspaces To Solve Problems

Function and form co-evolve. JL

Greg Satell reports in Digital Tonto:

Innovation strategies, when properly conceived, are focused on the particular type of problem they are meant to solve. So it makes sense that different strategies would require different workspace architectures. Apple, IBM, Google and Experian are very different companies which pursue very different types of problems. Each of these firms has designed workspaces with their innovation strategies — and the problems they seek to solve — in mind.

Why Amazon Is Partnering With Kohls To Manage Customer Returns

Both to maximize customer ease of doing business with Amazon - and to build alliances with bricks and mortar retailers in order to head up growing frustration with Amazon's commercial dominance. JL

Samantha Gowen reports in the San Jose Mercury News:

Amazon, in its relentless pursuit to be your one-stop e-commerce shop, has concocted a package return program with Kohl’s. The online titan has made strides this year to acquire or partner with brick-and-mortar stores, expanding its grip on the retail scene. Kohl’s, a discounter has maximized partnerships with name brands to grow its customer base.

The Race To Build A Computer Powerful Enough To Predict the Future

The ability to accurately analyze REALLY BIG data in real time - and then offer useful suggestions on what to do about it. JL

April Glaser reports in Slate:

Consumer laptops operate at gigascale speeds, which is 1 billion calculations per second. An exascale computer is a trillion times faster. The more powerful the computer, the more realistic the models it can create. There’s not currently enough computing power to model complex systems to uncover answers about climate change and scan social media in real time (which) could predict when someone is about to commit a crime. The US, Japan and China have initiatives to build exascale systems. No company that can afford to do this, and even a consortium of companies would not.

Sep 20, 2017

Why WeWork Thinks It's Worth $20 Billion

Not just a co-working play, but betting that data about office space - and how it is actually, optimally used - is as valuable, or more, than the real estate itself. JL

Jessi Hempel reports in Wired:

In the future, you won’t have your own desk, because your employer knows you use it 63% of the day. WeWork’s bet is it is amassing so much office space and studying how thousands of different businesses use it, it can position itself as the company with the most valuable knowledge about how jobs best get done. WeWork has been collecting data about how people work, where they are most productive, what they need to feel good, and how much space they require. “We help them make better decisions about what space to not have anymore."