In the last 5-6 years, machine learning has gone from ‘crazy idea from the 1980s’ to ‘software’. That has come with several waves of deployment and several waves of company creation, as we work out what do do with it. It’s the new SQL.Read More
We are in the middle of a wave of interesting new productivity software startups - there are dozens of companies that remix some combination of lists, tables, charts, tasks, notes, light-weight databases, forms, and some kind of collaboration, chat or information-sharing. All of these things are unbundling and rebundling spreadsheets, email and file shares. Instead of a flat grid of cells, a dumb list of files, and a dumb list of little text files (which is what email really is), we get some kind of richer canvas that mixes all of these together in ways that are native to the web and collaboration.
Then, we have another new wave of productivity company that addresses a particular profession and bundles all of the tasks that were spread across spreadsheets, email and file shares into some new structured flow. For example:
frame.io lets video professionals move version-tracking, reviewing and comments around their work from Google Sheets, email, and file shares (and Fedexed hard disks!) into a structured modern workflow
Everlaw (an a16z investment) lets lawyers move the discovery process (handling, reviewing and flagging hundreds of thousands of documents) from spreadsheets, email and file shares into a structured workflow, with machine learning analysis on top
And then, going one step further, Figma or Onshape (another a16z investment) pull the entire job (creating interfaces and 3D CAD respectively) from a stand-alone PC application that saves files to a file share into a new collaboration-first web application. (I would really like someone to do this for presentations)
Part of the point of both of these waves of company creation is that merely taking Excel or Word or Powerpoint and putting them onto the web may make the file share thing work a bit more smoothly, but doesn’t actually create a new workflow. If I’m a lawyer reviewing files for court, pasting the file names into a Google Sheet is probably better than pasting them into an Excel file on a network drive, but it’s still not as good as an actual tool.
In parallel to this, we have yet another set of company creation going on: the creation of vertical two-sided marketplaces, connecting some kind of service provider with their customers and providing tools and services around that. A16z has invested in Honor (home help) and Incredible (nursing), and there are lots of others, doing everything from truck drivers to oil field workers to dog walkers or gardeners. This might sound unrelated to ‘productivity’, but in fact all of these are unbundling and remixing spreadsheets, email and file shares as well, but from the opposite direction. If you want to book X, the old workflow is a spreadsheet or an old-world database, plus phone calls, faxes or email. And the way you solve this is not by moving it from Excel to Google sheets, any more than the way you solve version tracking of video edits or legal document review is to move it from Excel to Google sheets. You solve it with a dedicated workflow and tools. Incredible is Figma for booking nurses - or vice versa.
Viewed in that light, LinkedIn is part of both of these stories. LinkedIn tried to take the flat, dumb address book and turn it into both structured flow and a network of sorts. But by doing that for everyone, it has the same problem as a spreadsheet, file share or email - it’s a flat, lowest-common-denominator canvas that doesn’t capture the flows that many particular professions or tasks need. That’s the other side of the opportunity for Honor or Incredible: we unbundle LinkedIn, just as we unbundle Excel (my colleagues at a16z wrote about this here).
So, in 2010 Andrew Parker made a much-copied graphic showing how many startups were unbundling parts of Craigslist into dedicated vertical marketplaces. Today we’re also unbundling both LinkedIn and Excel, in the same ways and for the same reasons. There’s an old joke that every Unix function became an internet company - now every Craigslist section, or LinkedIn category, or Excel template, becomes a company as well. Depending on the problem, that might be a new collaboration canvas, or a new networked app, or a new network or marketplace, and you might switch from one form to the other. Github is a developer tool that also became a network - it became LinkedIn for developers.
But then, that isn’t static either. At a certain point, coolproductivityapp.io finds that people are a bit baffled by a flexible freeform canvas that lets them remix some combination of lists, tables, charts, tasks, notes, light-weight databases, forms and some collaboration, so you need to create some templates - maybe one for CRM, or events, or support tickets. And then… well, each of those becomes a company. The unbundler gets unbundled, or rebundled.
Meanwhile, a few years ago a consultant told me that for half of their jobs they told people using Excel to use a database, and for the other half they told people using a database to use Excel. There’s clearly a point in the life of any company where you should move from the list you made in a spreadsheet to the richer tools you can make in coolproductivityapp.io. But when that tool is managing a thousand people, you might want to move it into a dedicated service. After all, even Craigslist started as an actual email list and ended up moving to a database. But then, at a certain point, if that task is specific to your company and central to what you do, you might well end up unbundling Salesforce or SAP or whatever that vertical is and go back to the beginning.
Of course, this is the cycle of life of enterprise software. IBM mainframes bundled the adding machines you see Jack Lemmon using below, and also bundled up filing cabinets and telephones. SAP unbundled IBM. But I’d suggest there are two specific sets of things that are happening now.
First, every application category is getting rebuilt as a web application, allowing continuous development, deployment, version tracking and collaboration. As Frame.io (video!) and OnShape (3D CAD!) show, there’s almost no native PC application that can’t be rebuilt as a web app. In parallel, everything now has to be native to collaboration, and so the model of a binary file saved to a file share will generally go away over time (this could be done with a native PC app, but in practice generally won’t be). So, we have some generational changes, and that also tends to create new companies.
But second, and much more important - everyone is online now. The reason we’re looking at nursing or truck drivers or oil workers is that an entire generation now grew up after the web, and grew up with smartphones, and assumes without question that every part of their life can be done with a smartphone. In 1999 hiring ‘roughnecks’ in a mobile app would have sounded absurd - now it sounds absurd if you’re not. And that means that a lot of tasks will get shifted into software that were never really in software at all before.
Like Sky before it, Netflix is a television company using tech as a crowbar for market entry. The tech has to be good, but it’s still fundamentally a commodity, and all of the questions that matter are TV questions. The same applies to Tesla, and indeed to many other companies using software to enter other industries, especially D2C - what are the questions that matter?Read More
I sometimes think that if you could look in the safe behind Jeff Bezos’s desk, instead of the sports almanac from Back to the Future you’d find an Encyclopedia of Retail, written in maybe 1985. There would be Post-It notes on every page, and every one of those notes has been turned into a team or and maybe a product.Read More
Machine learning is the new centre of tech, and like all big new things there are issues. ‘AI bias’ is much-discussed right now: machine learning finds patterns but sometimes it finds the wrong one, and it can be hard to tell. This is a real concern, but it’s also manageable as long as we pay proper attention to it, and will probably look much like similar issues in previous waves of automation.Read More
Internet platforms are mechanical Turks - they can only understand things by finding a way to leverage vast numbers of humans. They’re distributed computers where all of us are the CPUs. How does that affect how we think about abuse, and how might machine learning change this?Read More
Apple’s talk about services got specific with a bunch of news subscription services. Most of them are sensible and worthy iteration, but the company still hasn’t explained exactly what it plans with its push into commissioning billions of dollars of premium TV (Spielberg! Oprah!). Maybe all of this is about trust: the old Apple promise was that you don't have to worry if the tech works, and the new promise is you don't have to worry if the tech is scamming you.Read More
Smart home today looks a lot like the world of kitchen gadgets a few generations ago - and so does machine learning. We have a bunch of cheap commodity components (DC motors! Cameras! Wifi chips! Voice recognition!) and we’re trying to work out how to bolt them together into things that makes sense. There are lots of experiments - some things will be the toasters or benders of the future, and some will be the electric can-opener.Read More
Facebook’s struggle with abusive behavior today looks a lot like Microsoft’s struggles with malware 20 years ago: people take advantage of an open platform, and you have to work out how far you can close the holes, how much you can scan for bad stuff, and whether you need to change the whole concept from the ground up. The answer to Microsoft’s problem wasn't Microsoft: we moved to the cloud and to secure OS models (iOS, ChromeOS). By pivoting to ‘privacy’, Facebook is trying to make the same move, but to do it itself.Read More
Machine learning means smartphones will (nearly) always take perfect pictures. But it also means they might understand what’s in the picture and why you took it. So what do they do with that? What does the discoverability and communication of AI look like, if you can answer lots of questions but might still be wrong?Read More
Amazon’s Alexa has been a huge, impressive and unexpected achievement. Amazon created a category from scratch and left both the AI leader Google and the device leader Apple scrambling in its wake. It’s now sold 100m units. So far, though, this success is pretty contingent - we do still have to ask what Amazon actually gains from this. What do consumers do with these devices that helps Amazon? What fundamental strategic benefit does it get? Amazon has put an end-point into tens of millions of homes - what does it do with it?Read More
What is 5G? Why do we care? How much faster does the pipe get? What can we do with a fatter pipe? How does this relate to VR? Cars? Broadband? What’s the killer app?
Really, unless you work in a few very narrow niches, you shouldn’t spend much time thinking about it.Read More
Machine learning is probably the most important fundamental trend in technology today. Since the foundation of machine learning is data - lots and lots of data - it’s quite common to hear that the concern that companies that already have lots of data will get even stronger. There is some truth to this, but in fairly narrow ways, and meanwhile ML is also seeing much diffusion of capability - there may be as much decentralization as centralization.Read More
Close to three quarters of all the adults on earth now have a smartphone, and most of the rest will get one in the next few years. However, the use of this connectivity is still only just beginning. Ecommerce is still only a small fraction of retail spending, and many other areas that will be transformed by software and the internet in the next decade or two have barely been touched. Global retail is perhaps $25 trillion dollars, after all.Read More
When Nokia people looked at the first iPhone, they saw a not-great phone with some cool features that they were going to build too, being produced at a small fraction of the volumes they were selling. They shrugged. “No 3G, and just look at the camera!”
When car company people look at a Tesla, they have the same reaction. The Nokia people were terribly, terribly wrong. Are the car people wrong?Read More
We're now four or five years into the current explosion of machine learning, and pretty much everyone has heard of it, and every big company is working on projects around ‘AI’. We know this is a Next Big Thing. I don't think, though, that we yet have a settled sense of quite what machine learning means - what it will mean for tech companies or for companies in the broader economy, how to think structurally about what new things it could enable, and what important problems it might actually be able to solve.Read More