Photo: Chris Liverani/Unsplash
Microsoft’s bid for the next generation of cloud startups

Good morning, and welcome to Protocol | Enterprise. In this Thursday's newsletter: Microsoft lowers the rent at its cloud app stores, you can't spell Dialpad without AI, and Google Cloud's Kelsey Hightower has a new role at the company.
(Was this email forwarded to you? Sign up here to get it in your inbox every week.)
Every year, Microsoft devotes one of the major events on its annual calendar to its partners, who collectively generate the lion's share of its commercial revenue. This week, that event — Microsoft Inspire — rolled around, and the headline news was the unveiling of a long-awaited cloud version of Windows for business customers.
But another announcement could have a much bigger impact on enterprise tech. Starting immediately, companies that list software in any of Microsoft's storefronts will pay a 3% listing fee when one of their customers buys that software. That's down from the 20% fee Microsoft charged a month ago, which it said was an "industry standard" fee for such transactions across cloud marketplaces.
Cutting listing fees could make Azure more attractive. Enterprise tech companies have been selling software on the cloud equivalents of Apple's App Store for years, but businesses hawking cloud infrastructure or application-development tools tend to debut on AWS, which has the largest number of infrastructure customers.
The new pricing terms will also have an impact beyond Azure: Microsoft's new fee structure will also apply to software publishers that want to sell apps for Microsoft Office, Outlook and Teams.
This is all clearly aimed at the next wave of cloud startups. Those startups — as always — should have a big edge over enterprise tech incumbents beholden to existing lines of business, and Microsoft wants a slice of their success.
— Tom Krazit
The confidential computing approach protects critical information by isolating sensitive data in a protected, hardware-based computing environment during processing. It allows governments to share the results of machine-learning inferencing on highly-protected or sensitive data sets without requiring them to share the data sets themselves.
---
Join Protocol's Jamie Condliffe for a conversation with Zoox's Ashu Rege, Qualcomm's Alex Vukotich and Luminar Technologies's Christoph Schroeder for a discussion on autonomous tracking and the technology that enables it.
July 21 at 9 a.m. PT / 12 p.m. ET Learn More
---
Can you hear me now? Dialpad is one of those aforementioned startups that thinks it can exploit new ways of working and collaborating with a fresh approach to communications backed by artificial intelligence. Protocol's David Pierce examined how Dialpad is evolving to meet the future of workplace communication and why more established companies like Zoom and Slack might want to pay attention.
Standards and practices: Web standards bodies are notoriously complex groups prone to infighting and intrigue, and the W3C's efforts to define privacy standards in browsers have been particularly fraught. This deep dive from Protocol's Issie Lapowsky into the technology and politics at the heart of the debate over the future of web privacy is well worth your time.
What was the first computer that got you excited about technology?
My first "computer" was a 1977 Radio Shack Tandy Science Fair "Digital Computer Kit." It wasn't an actual computer, but it did teach me about combinatorial logic. My first real programming was on a TRS-80 in 7th grade, and the first computer I owned was an Apple II+. I still remember "upgrading" it from 48K to 64K and migrating from cassette tape storage to a disk drive. These systems seem impossibly primitive today, but they revealed a whole new world of programming to me as a child — a world that has fascinated me ever since.
If Protocol gave you $1 billion to start a new enterprise tech company from scratch today, what would you do?
I'd go back to the drawing board on natural language AI. As exciting as ML advances have been, at the end of the day it's still not actual self-aware intelligence or learning, just clever parameter tuning. I think true silicon-based consciousness is within reach in our lifetimes, but it remains an undiscovered country that needs more explorers.
How can enterprise tech improve its current status around diversity, equity and inclusion?
There are a lot of obvious steps that we can, and must, take to improve inclusivity, starting with improved STEM education opportunities at the grade-school level. But I'll give you another, less conventional answer as well: Make software and software-based businesses easier to create and operate. Here's the rationale: The simpler software systems are to create, the more people who can create them, and the more diverse those makers and their ideas will be. Conversely, complexity breeds risk, and risk narrows opportunity by creating barriers to entry, where only the most privileged folks with Ivy League educations and corporate track records will receive the scarce funding and management opportunities.
Which enterprise tech legend motivates you the most?
Andy Jassy gave me my first opportunity to build a billion-dollar business, and I'm forever grateful for the opportunity. One of the amazing things he and others at AWS created was an innovation machine where people like me could come and turn ideas into businesses. I'm incredibly motivated by my personal experience there and by Amazon's concept of an innovation flywheel that rewards creativity and hard work with opportunity.
What will be the greatest challenge for enterprise tech over the coming decade?
Cross-party applications. Today, 80% of business data lives outside a company's four walls, with outsourcing, SaaS and increasingly complex supply chains driving that percentage ever higher. That means that classic IT approaches designed to address centralized applications, where the producer and consumer of data all work for the same company, are increasingly insufficient. Companies that get good at sharing information effectively with their suppliers, customers and business partners are the ones that will find transformative advantages over the next decade.
The confidential computing approach protects critical information by isolating sensitive data in a protected, hardware-based computing environment during processing. It allows governments to share the results of machine-learning inferencing on highly-protected or sensitive data sets without requiring them to share the data sets themselves.
Thanks for reading — see you Monday!
To give you the best possible experience, this site uses cookies. If you continue browsing. you accept our use of cookies. You can review our privacy policy to find out more about the cookies we use.