Month: April 2020

Our attention is finite – fundamentally, it’s a function of time. In an age of abundance of data, attention is capital. The most successful companies of our time look to accumulate and grow as much of it as possible: Google’s search and many other services, Facebook (with Instagram), Twitter, Spotify and of course TV news channels. Their business model is to resell our attention at a price to entities we have little to no control over.

These vast entities, most publicly listed companies, are Attention Merchants: we hand our attention daily to these entities. They aggregate and lend it out to advertisers for a price. We collect a fraction of the returns. Not as cash but as the product that is built with the cash. 

The problem is that it’s a closed loop – the product itself collects more attention and data. It’s pretty clear then that the primary incentive is not to make the product more useful for us but to capture every little bit of marginal attention. For example, Twitter’s Q4 FY19 investor fact sheet depicts people who use its service as faceless heads labelled ‘monetizable DAUs’, the primary company metric. This is not to single out Twitter – examine any quarterly result of any of these companies. The market rewards every quarterly increase in attention and its monetisation.

When that product has to do with information, with news, as nearly all of these do, the impact of this misalignment of interests is very real. 

For instance, rather than algorithms highlighting a finite number of tweets (or Facebook or Instagram posts…) that are relevant to me, all social media is now designed as an infinite river or stream of posts with an easy ‘pull to refresh’ (its inventor has regretted how addictive it has become). The algorithm may sort by both recency and relevance, but it is always infinite – the objective being to keep you on the site/app as long as possible. Other news and other content sites achieve this with ‘you may also be interested in’ widgets or outright loading other news stories as you approach the end of an article, again simulating a river. This means more consumption, less thinking.

The second-order consequence of this is the way online content is written. We have written about what happens when the audience for online writing is an algorithm instead of a human being:

This comment on Hacker News

I would add new ways of searching in Internet. If I searched for food recipes in the 2000s I would find independent blogs with some real local/family taste. Now I have a hundreds of results from click bait sites with the same commoditized recipes and the ugly blog with a good recipe deep in the long tail. We can say that we need improvements in the long tail when the tail deserves to move up (or to the left in a xy chart).

Internet search is a driver for the world economy, a tiny improvement would improve the life of entrepreneurs and their ecosystem beyond elite circles.

I had then written: Discoverability is going to become an increasingly large problem. Leaving aside content locks inside social media services, over and above what the commenter says, Google optimises for recency, page performance, mobile-friendliness. You can no longer filter search results by date range. Organic results are also pushed down the page by structured search results: top stories, news carousels, related search results, travel cards and more, all of which are dominated by publishers that support AMP.

Taken together, this ends up being biased against independent, casual, non-optimised but potentially interesting publishing, not to mention the loss of increasing amounts of old content that exists on the web but with no way to discover them any more.

This state of affairs creates a pervasive environment for distorted news. (part 3)

We need a new way of doing news.

Broadly, the current business model of production, distribution, discovery and consumption of news is built on hoarding attention to the detriment of education. This creates perverse incentives across the ecosystem to focus on a narrow range of short-term topics presented as entertainment. This is compounded by the fact that news organisations often align themselves to the world-view of their financier or benefactor.

But news also informs our world-view. Who we trust. How we vote. Who we make friends with. Where we move to. Where we spend and invest. This is especially true in a world that is going through as much change as quickly as today. It is far, far too important to be presented as a never-ending series of big fights. 

When you tease the problem apart, it becomes clear that it is not a question of one or two bad players, nor is it a conspiracy – those would be easy to tackle. The ecosystem itself is not incentivised to serve us – the reader, the viewer. 

At the same time it has never been more possible to set up, run and sustain an alternative way of informing the public. Technology has, to use a cliche, levelled the playing field. Most of the elements that make up 21st Century Media already exist. Now they need to be brought together for, and supported directly by, the reader.

Before we see what that looks like, we need to understand the problem better (part 2).

“Most of us kind of agree on the thrust of history. The key is to understand how we get there… The transition strategies are more important than understanding what the outcome state will be.”

– Sean Parker in his profile Agent of Disruption, 2011

We have discussed the problems with news before. And we have discussed the power of building an audience in the  constantly-online age. 

The confluence of these two can be summed up thus: attention is the capital of the 21st century. Media today – whether TV or on the internet – while always incentivised to hijack that attention for money, has now reached a point where it is either the direct or proximate cause of polarisation of society and its attendant problems by simplifying and packaging ‘takes’ on issues. 

This is especially unfortunate because we’re living in a time where disruption, no longer just a buzzword on a startup VC pitch deck, is changing industries, jobs and norms across the board and across countries. With the potential for great benefits it brings with questions of ethics and morality we must confront, and prospect of inequality we must resolve. 

I’d like to believe that today the first major step has been taken towards the way 21st century media should be.  

The journalist Matt Taibbi, writer of the excellent books Griftopia, The Divide and Insane Clown President among others, has gone direct: he is moving to a reader-paid email newsletter on the publishing platform Substack.

I am excited.

Not just by the move, but by his post justifying why. He gets it.

Both the new economics:

Compensation in news media traditionally involves a reporter working for a corporation or a wealthy patron, who ostensibly paid staff with revenue from advertising and subscriptions. This used to be necessary because delivering content was expensive and required additional labor: design, printing, distribution, marketing, etc. 

Distribution is instant now, design can be automated, and there are no printing costs. The logical endgame is cutting out middle steps and having journalists work directly for readers. 

and the more important imperatives:

The media business as constructed is expert at mass-generating binary streams of hot takes and talking points, and selling ads to a public engaged by them. It’s great business: cable profits have soared. But it’s a lousy system for getting to the bottom of difficult subjects, and boy do we have a lot of those to deal with

There have been independent political bloggers before, most notably Andrew Sullivan of the Daily Dish, which was also paid-subscriber model. But that was seven years and a lifetime ago. I hope this is what launches the new new media.

This 2019 study (PDF) of citizens’ data literacy in the UK by the University of Liverpool, Glasgow University and others has one curious conclusion: a profile they call ‘social and media users’

On the surface they resemble what you would typically call digital natives: young, who spend large amounts of time online and share lots of content. 

But they are different in that they are significantly less aware of the source and impact of the information they receive and share. They are also likely to be less educated than other profiles. According to the report:

‘Social and media users’ have almost as limited an awareness of the use of data by platforms as Limited users. At the same time they have the least concern about data sharing and the least critical position on the data sharing practices of platforms. 

Ironically, they also still do not trust content they find in any media – but they are more likely than other groups to trust content shared by friends. Given that this group (17% of users) consists mainly of young people, with lower educational attainment from lower income households, we are concerned that they will remain disadvantaged in their data literacy into later life. 

This result – along with similar ones elsewhere in the literature – undermines the idea of the “digital native”. 

Other details about this profile that stood out to me:

• 95% reported to make no effort to view news website which are different political perspectives than their own
• Just 30% looked online to verify information during a conversation with friends or family
• Only 8% encouraged or showed others how to fact-check things online, for example by conducting other searches
• And in a dichotomy, 85% indicate that it is not acceptable to track their online behaviour over time but 50% believe that there is no point in changing privacy settings because companies will be able to get around these settings anyway

These are a passive group that is ripe to be influenced with misinformation online, such as the campaign run on Facebook during Brexit. Their consume-and-share behaviour puts not just people like them at risk but other groups too.

A friend asked yesterday whether most people even know of the iPad Pro’s and iPadOS’ capabilities, leave alone care. So. A short thread on what I think Apple’s thinking is regarding the iPad Pro buyer segment:

The iPad Pro is specifically targeted at that narrow segment of people who already know, or educate themselves, on what the device’s capabilities are. Most others don’t care – and for them, there exists the regular iPad.

Now the regular iPad gets some of iPadOS’s software capabilities for free, such as pointer support. But in terms of raw performance and hardware, it lags. It’s unlikely to get multiple cameras, LIDAR, Smart Connector, the speaker system, even ProMotion.

But that keeps prices down, and that’s key. The causal buyer will look at price first and features second (they are already sold on the core iPad proposition). The vanilla iPad is a killer everyperson tablet at its price.

The professional, however, is going to want a device with whole range of features and capabilities because their iPad use cases form the long tail. And for that power and functionality, they are willing to pay lots more:

The question I think Apple wants you to ask yourself is Do I *rely* on my iPad for work? For one buyer segment, the first time they ask themselves that is when they will realise they have outgrown – instantly – their regular iPad. 

Apple’s ‘What’s a computer’ ad campaign earlier and ‘Your next computer is not a computer’ campaign now is to get that segment to ask themselves that question:

So. The non-Pro versions are what the vast majority will buy. But because it has more gee-whiz, the Pro is what will get the most press. And that press creates a halo effect around Apple, deepening that sense of desirability. 

End note: Apple has had a long history of regular and pro variants of its hardware lines. You could even say it went all the way back to the Mac (simple, closed) and the Apple II (complex, extensible). The iBook and the PowerBook. The MacBook and the MacBook Pro. Today the iMac and the iMac Pro.

The iPhone and iPhone Pro difference, though, I think is somewhat facile, merely branding. In fact the regular iPhone (and previously the iPhone XR) are arguably better for heavy everyday ‘pro’ use – not only do they have nearly everything the Pro phones have, they have longer battery life because it doesn’t have to drive the more power-hungry OLED screen. I wonder if that hurts the Pro branding for other product lines where it it more meaningful.

In Federico Viticci’s excellent long-form blog post about the iPad Pro as a modular computer is this section about accessing his Mac (a Mac Mini):

With my Mac mini running in the background, I can open Screens on the iPad Pro, which instantly logs me into macOS with my credentials. Here’s why this setup is incredible: Screens for iPad supports full-screen output on external displays (more in the next section), which means I can interact with a full-screen macOS UI on the UltraFine display that is actually being transmitted from an iPad app over USB-C. In the latest version of Screens for iPad, I can use the Magic Trackpad to click-and-drag macOS windows, right-click to open contextual menus, and otherwise use the native macOS pointer from my iPad without even seeing the iPadOS pointer on my external display. It’s a mind-bending setup, but it works beautifully – you’d be forgiven if you looked at the photo below and thought I was using macOS and the iPad Pro next to each other. In reality, that’s just my iPad Pro running Screens in external display mode along with a Magic Trackpad 2… Effectively, this is macOS as an app.

– Modular Computer: iPad Pro as a Tablet, Laptop, and Desktop Workstation

This is fantastic. And a great way of putting it. The iPad Pro’s performance makes it possible to interact with a remote machine – graphically. And the software support both within the Screens app and now within iOS 13 make it possible to emulate Mac OS gestures via inputs connected to the iPad Pro.

My needs don’t require a full-fledged Mac OS desktop, but I do require a UNIX setup occasionally. In that case I use the excellent Termius app to SSH into my Raspberry Pi that I have physically attached to my router and mainly runs Pi-hole.

Termius has a built-in SFTP client, but I’m not sure I want to run an FTP server on the Pi. So I use Resilio Sync to transfer any files to/from the iPad. This works for the most part – Resilio Sync on the iPad even has a file provider so it shows up as a destination in Files and file browsers views, but the app needs to be open to actually sync. I can live with it, but it’s suboptimal, especially when transferring large files.

CS Music: formerly Cesium: iOS’s music app used to be really good. No more. Hence.

Overcast: iOS’ podcasts app used to be the best. Again, no more

Readder: Main, fun Reddit account

Apollo: Account with fewer, more focused subscriptions

Blockquote: plaintext editor with great iCloud Drive support

Files: files app with great iCloud Drive support

Sheets: personal and work spreadsheets. Many populated by IFTTT recipes

Health: tracking water intake (populated by an iOS Shortcut), meditation (“Mindful Minutes”, also populated by Shortcut), heart rate (populated by Soul Solver via a Fitbit).

Evernote: research material for projects

Microsoft Todo: main todo app. Does lists better than iOS Reminders

Launch Centre Pro: Launches apps and Shortcuts manually but based on time and geofencing. I use it multiple times every day and almost feels part of the OS.

Cloudflare has an interesting microsite that emphasizes how the open nature of the Internet has made it the global connective layer we are plugged into during this period of global isolation. Worth a read. 

The stage was set for an always-on, always-connected future. A demanding world where every user would choose how they use the network. And the network was almost ready. In the meantime, home and office computer users were forced to dial into private services like AOL, Prodigy and CompuServe. These closed worlds contrasted sharply with the open Internet growing quietly in the background.

The Internet we know today was dreamt in the 1960s, debugged in the 1970s and built in the 1980s. Thirty years of work to create a reliable, flexible, inexpensive, global network.

As 1990 dawned it was ready for all of us. Thirty more years later, in 2020, it binds us together.

In the next few years, during the next global crisis, we will debate issues like net neutrality, like data ownership, like censorship, because there will be issues of not just the economy but also of life and death riding on policy around these issues. 

Case in point: during the 2018 California fires, Verizon famously throttled the fire departments’ speeds because they were out of data. There is an argument that they should have been compelled to provide the best possible bandwidth to an emergency service, against the principles of neutrality. 

Similarly, you could make the case that bandwidth on Zoom (or Skype or other video collaboration services) be reserved first for governments, medical and healthcare personnnel, other essential services and then other (possibly paying) businesses. Again, against the principles of neutrality.

Censorship examples are easier to find and make. The last twenty years was using terrorism as the bogeyman, the next couple of years will likely be using that of a public health crisis. India has a problem of misinformation spread by people over Whatsapp. During the current global pandemic, it’s easy to spread panic via rumors and easy to risk health by touting suprious cures. There is a case to be made for censoring content on these private messaging platforms. 

We don’t have the answers, but we will need to confront these questions in the years right ahead.

This important article in the Indian publication Medianama on the pandemic bringing to the fore the importance of the Internet, and how it should be treated on par with other infrastructure, like say electricity.

It recognises that Internet service and e-commerce have been listed as essential services for the national lockdown, and in fact uses this to argue that this should be formalised in “broader legal and institutional frameworks”

It talks about delivery of government services:

The present situation also threatens to exacerbate India’s ‘digital divide’, denying access to public goods to those without internet access. The delivery of government services, in particular, is increasingly dependent on internet connectivity at the household level. The Aadhaar project relies to a large extent on internet connectivity at the Point-of-Sale level, which routinely fails. 

And the hardships faced by healthcare workers in Kashmir during the restriction of Internet access in the area.

The article also refers to e-commerce delivery workers. This is what interests me the most. Against all odds, a massive country with a messy transportation system, a severe lack of urban planning and universal addressing, and a sharp urban-rural divide has built a nation-wide delivery network via its e-commerce firms and their fulfilment partners.

If viewed and operated as infrastructure, even if owned and managed privately, this network can be used to deliver all manner of essential goods accurately, and even collect payment. It can even be used to transport personnel, say medical personnel, especially for the last mile. And not just during a pandemic or similar state of national emergency. India could rethink healthcare, education and financial inclusion with such an infrastructure. India Post Bank and Paytm Payments Bank already tech and a widespread delivery network to offer their banking and investment services across the country. But it could make entirely new things possible – such as a medical trial, which involves delivery and collection with rigorous tracking, at scale for specific categories of people, say ethnicities.

Never letting a crisis go waste is a cliche, but re-framing some assets as infrastructure could truly make this happen.