Category Archives: Startup general interest

Software doing the work of cameramen

By | Startup general interest | One Comment

“Software is eating the world” and “robots are taking all the jobs” are much used phrases these days. That’s because they are regularly proven to be true in new and surprising ways.

I was surprised earlier today when I read this excerpt from a Fast Company interview with Chris Anderson, founder of 3D Robotics:

“The first phase of our little adventure was getting robots to fly. That was super hard, but we got there,” he says. “The next phase was putting cameras on them, and stabilizing with a gimbal. That was pretty hard, but we got there, too.” What we’re missing, he says, is “the aesthetics of a good shot.”

Solo, in concert with GoPro, is designed to deliver that perfect shot, with very little technical skill on the part of the pilot. “There are these well-established Hollywood conventions about what makes a great shot; they have this combination of classic framing and paths, which are typically done by teams of professionals,” he says. “We turned all that into software.”

The first paragraph is a cool description of the challenges in getting drone mounted cameras to work well. The second paragraph is the best though. 3D Robotics have captured the art and craft of Hollywood camera professionals in software. If it works as well as Anderson implies then it could be the sort of highly visible development that helps people to viscerally understand that change in employment patterns is coming fast and that we should start preparing.

Government and startups: It’s not them vs us

By | Startup general interest | One Comment


There’s a widespread notion in the tech world that governments get in the way of innovation. The idea that slow moving, self-interested bureaucracies make it harder for nimble entrepreneurs to change the world by raising taxes and creating regulations is an easy narrative – but it’s wrongheaded.

The problem starts with a mistaken view that the public and private sector are separate entities in conflict with one another. There is no point where one stops and other starts. Rather they bleed into one another. Society needs a range of services to function and some are best provided by government, some by the private sector and many by a blend of the two. Transport infrastructure and defence are examples of ‘public goods’ that are best provided by government. The prison service and the health service are amongst those best provided by a blend of government and the private sector. Finally, food and clothing are on the long list of things best provided by the public sector.

In different times and in different countries there have been a multitude of experiments where services have been moved between the public and private sector, but these three categories have always been present, from black market economies in communist Russia to transport infrastructure in the USA.

The chart above shows how theory translates into practice. It’s from a Techcrunch article urging startups to partner with government rather than fight it. I couldn’t agree more. It’s what society needs, and as we can see from the frequency and size of the green bars above, it’s a good way to make money.

Choosing people over algorithms

By | Startup general interest | 4 Comments

It seems to me there might be an emerging trend back towards human curation. LinkedIn’s new Pulse App features human curation, the same is true for Apple’s forthcoming news app, and Apple’s new music service similarly makes a big deal about it’s human editors, including former BBC Radio 1 favourite Zane Lowe.

Two of these examples are from Apple, making it early to generalise, but there’s definitely a strong meme that humans are better than algorithms when it comes to recommending content.

I wonder three things though:

  • whether large companies prefer human curation because it leverages one of their strengths vis a vis startups – capital
  • if this is more about positioning than substance – human curation sounds better to most consumers than recommendations from a machine
  • whether the real answer is a combination – human curation augmented via AI – we’ve made one investment on that basis and are considering another one now

Let’s design the future of capitalism

By | Startup general interest | 7 Comments

As regular readers will know I’m optimistic about the future. I think there’s a strong chance that advances in technology will bring us cheap and abundant energy, machines that can do most of our work for us and medicine that delivers longer, happier and healthier lives. But, as I’ve also written before, the path between here and there will most likely be very rocky. Automation will replace jobs at an increasing pace over the next decade or two and without radical change wealth inequality will skyrocket to dangerous levels and existing welfare structures will collapse.

My kids will enter the workforce in 10-15 years and I’m worried by what they will find.

There’s a counter argument to this, most vocally espoused by Marc Andreessen, which says that every time technology replaces jobs the capitalist system finds new work for people to do. We’ve seen that movie play out multiple times over the 200 years since capitalism rose to prominence in the UK’s industrial revolution and we should expect to see it again.

The only thing I know for sure is that there are no certainties, and Andreessen might be right, but I think it’s more likely that this time it will be different. If I’m right it will either be because the destruction of jobs will happen much faster this time and the job creation won’t come close to keeping pace, or because automation will take the new jobs too and there will be a permanently lower requirement for human labour going forward.

In both these ‘it’s different’ scenarios we will need more income redistribution to fund bigger and better safety nets and radically better retraining and back-to-work programmes. Otherwise we will end up with a large permanently unemployed underclass and riots on the streets. A universal basic income is one solution that’s being increasingly widely touted.

Right now it’s still hard for most people to believe that we are headed to a post-scarcity world and think it’s a waste of time thinking seriously about how we navigate from here to there. The common reaction to pending automation is to fear job losses and robot overlords, think briefly about restricting technology development, decide that’s futile and then put off change for a few years because the problems aren’t imminent and the solutions are hard. The point they’re missing is that most of the pertinent technologies are developing on exponential curves – change will come slowly and then it will come FAST.

Capitalism is a system designed to optimise the distribution of scarce resources. That’s what money does. If we are entering a post-scarcity world then almost by definition we will need a new system. If we are entering a period of difficult adjustment then keeping equality of wealth and opportunity within reasonable bounds will be a difficult global challenge. Either way, we should take the opportunity to design our future system rather than simply let it happen to us.

Designing our future system requires thinking through where we collectively stand on acceptable levels of wealth inequality and how much we support the right to work. That’s worth doing even if I’m wrong and Andreessen turns out to be right.

For further reading see Vivek Wadhwa’s recent post on Venturebeat.

Retail’s future is to solve the paradox of choice

By | Startup general interest | No Comments

The primary function of retail used to be access/distribution. In the vast majority of cases consumers had to visit high street shops to get products and manufacturers went through physical retail to reach customers.

One of the big promises of the web is that brands/manufacturers will form direct relationships with consumers. They can build their own showrooms and take orders direct over the web capturing the retail margin for themselves. So far it’s mainly startup manufacturers that have gone this route – Bonobos and Warby Parker are famous examples from the US and LostMy.Name and Spoke from our portfolio are blazing a similar trail over here. Established brands are now moving slowly in this direction – but they need new capabilities and have to worry about conflict with the retail partners that drive the vast majority of their sales. Most established brands that now sell direct still only have a fraction of their product range available on their own sites and pricing is often above what can be found in retail.

So retail’s raison d’etre is losing relevance.

But consumers are faced with a new problem – too much choice. The web makes all the products in the world available, and a wealth of research has shown that whilst we say we want choice, too much of it makes us less happy and results in fewer purchases. This is the Paradox of Choice.

The new opportunity for retail, therefore, is to solve the paradox of choice. Give consumers access to all the inventory there is and then help them to a decision. At it’s heart new retail will be about personalisation and recommendation, making smart use of data and artificial intelligence to guide people to purchase decisions. Conversational interfaces will be an important secondary part of the mix as each of us will effectively have to programme the services we use and very few of us have the patience to learn syntax and commands for new apps.

These new retailers will sit between brands/manufacturers and consumers. They will handle less stock, have fewer staff and take less margin than traditional retail, but they could/should be much more profitable. Thread, Stylect and Top10 are good examples of these ‘new retailers’ that we’ve invested in.

Matching product solutions with problem spaces

By | Startup general interest, Uncategorized | One Comment

We held our monthly office hours yesterday morning, meeting with twenty idea stage founders over a period of two hours. It’s a a high octane set of fast paced meetings that I look forward to very much. It’s great meeting lots of great people and learning about lots of great ideas so quickly.

The last one I met had identified a large and fast growing market riven with inefficiency. The more we delved into it the more excited I got. Big markets with lots of opportunities for improving things are great places to start new companies. Then, when we got to discussing his company’s product/service, I knew that he wasn’t yet at the point where we could have a serious investment discussion. He knew that existing products in his market were of inconsistent quality and that consumers go to extreme lengths to buy them anyway (often gathering in groups to buy wholesale on AliExpress), but was only just coming to the conclusion he wanted to build a product company and had a feeling that the best opportunity lay at the premium end of the market but couldn’t articulate what his product differentiation would be or how he would price it.

That’s fine. We will keep talking and help him through that process, but right now he has identified a great problem space but hasn’t matched it with a compelling product solution.

That contrasts with another company we are talking with where the founder started selling a product to his friends and family because he loves delivering it and they love buying it. In this case the product solution is already defined in great detail and we have been working with him to work out how big the opportunity is.

In other words he has built a compelling product solution and it’s unclear whether it’s in a great problem space.

The best business opportunities match great problem spaces with compelling product solutions. Most businesses start with one and then work to find the other. “Technology looking for a problem to solve” is a cliched way of describing interesting solutions with poor problem spaces, and perhaps more common but lacking a common descriptor is companies in interesting problem spaces constantly ideating in search of a product people want.

Entrepreneurs who know their problem space but not their solution should, with discipline, be able to ideate, research and iterate their way to a compelling product. Entrepreneurs with a compelling solution are in a more binary situation – either their problem space is interesting enough or it isn’t.


Designer babies and other moral dilemmas

By | Startup general interest | No Comments

People have talked about designer babies for a long time, but I just read for the first time about a clinic openly offering genetic screening of embryos. You can pay $18,400 to screen for gender and they plan to soon offer eye colour, hair colour and other traits.

That’s illegal in most countries, but (somewhat surprisingly) it’s allowed in the US.

Genetic screening has reached this point now because technology developments are improving capabilities and driving prices down. That’s a phenomena that’s happening across a broad swathe of areas at an accelerating rate that will throw up moral dilemmas we have to grapple with.

  • Designer babies
  • Human augmentation with robotics
  • Human genetic enhancement
  • Cloning
  • Autonomous robots taking jobs (self-driving cars will be the first)
  • Autonomous robots making ethical decisions – e.g. a self-driving car choosing between hurting it’s passenger and a pedestrian in a crash situation
  • Artificial intelligence advising on legal matters, and eventually de facto making decisions
  • Artificial intelligence advising on policy matters, and eventually making de facto decisions

Coming to the right answers on these questions will require solid thinking and strong, far-sighted leadership. The default reaction of many people, politicians included, will be to reject, suppress and deny. That risks pushing developments and scientists underground and/or to less restrictive geographies where they will still flourish, but in a less controlled manner. And then they will come back.

And this isn’t far away now. In the next 10-15 years most of these technologies will reach a price point that makes them widely affordable.

We have interesting times ahead.

Spreadsheets in legal documents

By | Startup general interest | One Comment

I have spent a considerable part of today translating the equalisation clauses in our Limited Partnership Agreement into formulas in a spreadsheet. Equalisation payments are made when new investors join a Limited Partnership and they pass from new investors to old investors to compensate the old investors for the additional risk they took by investing first and/or to adjust for changes in the value of the fund’s investments in the period between when the old and new investors invested. They are dependent on a number of inter-related variables which makes them hard to calculate.

That was time I didn’t expect to spend on this task and time I didn’t have. And we still have to get all parties to understand and buy into our spreadsheet so it isn’t over yet.

I’ve been here before too, when implementing anti-dilution clauses after down-rounds.

It shouldn’t be necessary.

Rather than imperfectly describe formulas in legal documents we should be able to capture them precisely in a Google Sheet or Excel and append the file to the legal document. Then we would have definitive reference calculations that everybody could use saving time and eliminating scope for disagreement over interpretation.

I believe my former employer Operis have started doing this in the project finance arena, attaching complex models of entire projects to legal contracts which are referred to in clauses determining how payments change under different circumstances.

In time the Blockchain might be a better solution to this problem, but spreadsheets would work today without further development of software.

“Sharing economy” is running its course as a useful term

By | Startup general interest | 8 Comments

The meaning of the term ‘sharing economy’ has always been a bit vague for me. On the one hand it includes genuine sharing ideas like Airbnb whilst on the other it includes labour markets like TaskRabbit and Homejoy, and then more recently Venturebeat lumped in companies like Wework and Transferwise (although they called it the ‘collaborative economy’).

Still, the label was useful for PR purposes, and for building a positive company image with investors, regulators and the public.

That may now be changing. I’ve just read Sarah Lacy’s Let’s face it: Uber IS the sharing economy which makes the point that Uber dominates the sector, and The “Sharing Economy” is the Problem which argues that Uber and other sharing economy labour marketplaces are using a novel corporate structure to exploit workers.

Being labelled as ‘sharing economy’ is starting to carry more negative connotations than positive. Hopefully we can move to something clearer.

UPDATE: Alan Patrick pointed me to a Grist article which makes a more emotive claims against Uber and others:

The sharing economy is a nice way for rapacious capitalists to monetize the desperation of people in the post-crisis economy while sounding generous, and to evoke a fantasy of community in an atomized population. …

[I]t sees us all as micro-entrepreneurs fending for ourselves in a hostile world. … You may lack health insurance, sick days, and a pension plan, but you’re in control.

The lock of traditional TV distribution is weakening

By | Startup general interest | 3 Comments

Earlier in the week I wrote that the time for TV and film distribution startups might finally be upon us, arguing that radical reductions in the cost of production are a force for change that could break the legacy arrangements that have held for so long and usher in an era of direct connection between artist and consumer.

This chart shows that TV viewing is declining fast amongst American teens, weakening one of the key points of legacy control and creating more space for new distribution startups to build an audience. I haven’t seen equivalent stats for the UK, but I imagine they would be similar.


Get Social

Blog Newsletter Sign Up

Enter your Email:
Preview | Powered by FeedBlitz