Game Design, Programming and running a one-man games business…

The other kind of silicon valley tech bubble

We often read talk about how there is a ‘tech bubble‘ or more accurately ‘another tech bubble’. People with long memories can recall the insanity of, and then look at current valuations for airbnb and uber, and wonder if there is another day of reckoning coming. Maybe there is, or isn’t, I honestly have no idea. There is however, another ‘bubble’ issue with silicon valley, and in some ways its both more worrying (for what it says about society) and more destructive (for whom it hurts). To explain the bubble, I need to talk about my car, and its autopilot features.

I am stupidly fortunate enough to drive a Tesla Model S, with autopilot. It is AWESOME. Its by far the best car I have ever owned. I love Tesla, I have Tesla stock, I believe in the company, I have a Tesla T-shirt, I’ve read that book about Elon Musk, I am a Tesla fanboy. Autopilot is amazing, and cool, and awesome, and worth the money. Now let me tell you why it is completely oversold, overhyped and rubbish.


Autopilot does exactly what its advertised to do. On a motorway (highway to you Americans), its basically a self-driving car. It is pretty flawless at staying in lane, steering, and changing lanes, and the ‘Traffic aware cruise control’ is awesome. On smaller, but good quality wide A-roads, its also amazing. Its very, very good at what it does. The trouble is, all the things it really sucks at are very interestingly all the things you don’t get much of in Silicon Valley.

Autopilot is bad, even maybe a bit dangerous if one side of the road is unmarked with lines, has no kerb, and a hedge. It hugs the side of the road way too much in that case. Its not very good in heavy rain, where you have dirty roads and muddy roads and the lines are obscured or just not even there. It doesn’t like cars parked on the roadside. It is absolutely useless / dangerous if it encounters a roundabout (mini or otherwise).  None of these are surprising to me, none of these are a problem. I drive with autpilot on maybe 20% of the time (at most). (By the way, I live in a tiny tiny village in rural England.)


So far, so good. The problem is, lots of the tech/money people in silicon valley seem to think driverless cars are imminent. No Fucking Way. A car that does 95% of your motorway driving? Sure. A car that maybe does 90% of your driving in general? Sure. But a car that does 100% of your driving and you can read a book? Not now, not soon, not for a lot longer than people think. The trouble is, if you live in Silicon valley, you commute from your home with a garage, drive along highways and wide open US roads in California sun, never encounter a roundabout, never get stuck behind a horse, never see a road that has the remains of straw bales scattered all over it to cover horse crap…never encounter any of the 1,000 other ‘anomalies’ that I see on my roads every day…then sure! Self-driving is imminent. And those hedge fund managers who live in ‘grid-layout’ new york will agree with you too. The car industry as it was is DEAD. All cars are about to become robots.

Fucking hysterical.

The most dangerous thing in the world is to think that everyone lives like you, thinks like you, and wants what you want. That way, you start to disregard whole areas of thought, whole groups of people, and become insular, closed minded and prejudiced. The vast majority of non tech people I know do not give a DAMN about self-driving cars. They want cheaper rent, more stable jobs, better pay. They are happy if they can afford *any* car. The idea of lusting after a self-driving one is a laughable past-time for the super-rich.

Leave this problem with cars and it doesn’t matter. I think it leads to over-optimistic tech stock valuations, but that’s no big deal in the grand scheme. The problem is the ‘driverless cars are here’ cries are a symptom of a wider problem. Most of the people with big financial clout are living in a bubble, where the only concerns they understand are the concerns of *people like them*.

Here is a shock: My phone is cheap enough, light enough, and has enough features. My laptop is thin enough, light enough, and has enough features. By the way so does my car, so does this desktop PC, so does practically everything I own. I have some fucking virtual reality goggles for crying out loud. if you ask me what is *missing* from my life, I’d have to start dreaming up some really crazy stuff. maybe a self-filling voice-activated kettle?  errr maybe it would be nice if my TV was voice activated. errr…

And this is dumb as hell because there are VAST swathes of people out there who can rattle off their top needs with no problems. Ask a Syrian refugee, or someone in a country with no running water or mains electricity what their needs are. or better still, don’t even go that far, just walk outside your luxury office with its canteen with pastry chef and neck massages for all staff, into the streets of San Francisco and ask someone homeless what they want. I doubt they will start talking about how they hope the iphone 7 has a new headphone connector. Its not going to make their list.


I get the economic argument. Homeless dude and Syrian dude have no money, whereas I do. So you try and sell me a new phone to replace last weeks phone. I understand the maths. I understand that trying to develop a business model where you can provide goods for people with very little money is REALLY hard. But the tech elite keep telling us how clever they are. If so, you guys can do it. You just need to step outside your silicon valley bubble and take a look in the real world. Thankfully, and ironically given this blog post, the one guy who seems to ‘get it’ is actually Elon Musk. The Model S was a stepping stone to the model 3, and I fully expect the cars to get cheaper and cheaper until *shock horror* they are available to ordinary people. My question is… where is the Elon Musk for food, for housing? for education? It seems like the ‘real world’ problems of access to education, clean water, food, shelter, are ‘too dull’ or maybe just too boring for the tech elite to bother with. That doesn’t need to be the case, they just need a little imagination. And if the only thing they can possible invent is a slightly thinner, slightly faster phone? then how about just paying the fair share of corporate tax, and let the government use it to do the ‘dull’ work?

By the way, if you do happen to be a tech-elite type, and feel bad reading this, you can build a whole school in cameroon for about $25,000. It’s easy.


13 thoughts on The other kind of silicon valley tech bubble

  1. My problem with the Uber/Airbnb bubble is that the dot-com bubble at least was based around a genuine technological innovation – delivering goods and services using the Internet. Uber and Airbnb on the other hand, are largely based around circumventing if not outright breaking the law – taxi regulations and zoning and hospitality laws respectively. Those giant valuations are based on the assumptions that the states and cities will not start enforcing the laws more vigorously and that the local laws will be just as easy to circumvent in other parts of the world, where the companies will inevitably attempt to expand.

  2. While I agree with your assessment that self-driving cars are probably not around the corner, and that a lot of tech developed in SV only works in SV, I feel like this post is mostly a rant without much substance. Have you tried to look for companies that work with life-improving tech for poor people? This post would be much improved with some statistics.

    I don’t know how it is in SV, but in my own startup hub (Stockholm), a lot of the entrepreneurs I run into are working on these “dull” problems with quite clever solutions.

    Also, I cant help but notice the irony of this complaint coming from a video game developer. People might not need thinner phones but they surely don’t need another game either. Why aren’t you working on stuff that help poor people instead?

    Sorry if I sound harsh. Love your blog usually :)

    1. I’m better at coding games than I am at building houses, which is why I took some of my games profits and built a school in Africa with it. Also, I don’t spend half my time making corporate videos about how I’m ‘making the world better’ like companies in SV do :D

  3. Great post! Very thought provoking.

    Most of the companies working on self-driving cars (Google, Uber, even Ford and Tesla) are either intimating or outright saying that they intend to focus initially on rideshare use in urban centers. Elon Musk aside, most don’t seem interested in selling autonomous cars to private individuals that can drive them from the Cotswolds to Heathrow and then drive themselves home again; those problems, as you point out, have a long tail, and the solutions scale poorly. Instead, they want to tackle dense urban centers, which have wide avenues, cleanly marked roads, are very well mapped, and benefit from lots of potential users. If you can get a car good enough to take you from any bus stop in Los Angeles to any other, even if it can’t handle a residential street, you’ll be within two blocks walking distance from almost all Los Angelenos and the places they work, and by skipping all the stops and transfers you’ll be a lot faster than a bus; fast enough to make an unworkable 2 hour commute into a survivable 40 minute one. As a bonus, roads with bus stops on them tend to be wide and straightforward to drive (they have to be, or they wouldn’t be able to accommodate busses).

    “Replacing busses” isn’t as impressive as perfecting robotic driving, but it could be a lot more socially impactful. There are a few cities in the world with public transit ubiquitous enough, fast enough, and reliable enough that most people don’t need to own a car, but there are a lot more where they do, and unless you’re quite well off car ownership is expensive. In the US, the average household spends about 18% of their after-tax income on transportation costs ( only slightly less than housing, more than food, much more than education. Most of that is the ownership, insurance, and fuel costs of owning a car that spends most of its time parked. Cutting that number in half for people, by giving them a cheaper practical commuting option, could be transformative for a lot of lives. It’s hard to imagine any other comparable tech-driven disruption in the cost of housing or food on the horizon (though, in a way, self-driving transit could help with both: giving people more manageable commutes widens their housing options, and reducing long-haul overland shipping cost would reduce one of the major drivers of food cost).

    There are downsides: a lot of those “stable jobs” people need come from selling and maintaining city dweller’s cars, gas, insurance, etc. But I still think a level of autonomous car that could take over lots of meaningful urban transportation is pretty close, and has the potential to affect the lives of far more (and more diverse) people than Palo Alto millionaires; in fact, the “bubble” class, for whom transportation cost is a rounding error, is likely to be affected soonest, but least meaningfully.

    1. Very well put. Its true that there are definitely relatively closed-loop environments in city centers that could benefit from self driving cars LONG before they become ubiquitous. Even then, this is more likely to happen in the US long before other countries due to the relative simplicity of US city road layouts and the width of US roads.
      Central London, for example, is a self-driving cars nightmare, in comparison with Los Angeles or San Francisco.

      I’m not *against* people innovating in this area, but it just seems crazy for a company like LYFT (for example) to try to topple Uber, when there are so many other areas ripe for innovation and disruption where the tech billions are not being poured.

  4. As a former AI programmer this self-driving car bubble does seem awfully familiar from the last time people got carried away.

    The last time natural language communication was a biggie people magically translated simple tricks (eg. Eliza) into ‘Computers will soon be able to have a conversation with us’. I assume they’re still waiting. Hubert Dreyfus’ ‘What Computers Still Can’t Do’ (1972, revised 1992), is still valid in its criticism of rampant over-optimism about what AI can do even if some of the examples are a bit dated.

    The advantage of fully automated vehicles however is massive in terms of productivity improvement, which as an economist I’m sure you’ll appreciate.
    In the longer term the cost of transport would collapse, making everything cheaper for everyone. (See Paul Mason on the liberating effects of automation for normal people, and the underlying value of the labour-cost of items).
    This probably explains the hopes and magical thinking – practicality be damned – about AI vehicles. Who cares if you can’t actually do it? It looks convincing and it’d be great if it DID work.

    I really should get a list together of situations that AI cars have no hope of dealing with. I hadn’t realised they were so bad on merely poorly marked roads.
    These are some situations I’ve personally experienced:
    – The Magic Roundabout, as above
    – The reasonably major road going past my house suddenly changing to a one-way street overnight.
    – Devon.
    – Broken traffic lights, which were stuck on red at one end.
    – Driving the wrong way on a motorway and leaving via an entrance ramp (Motorway was blocked by an accident)
    – Horses

  5. There are firms that are making food cheaper through technology such as genetic engineering, including Monsanto. But people have convinced themselves it’s evil and governments (especially in Europe) have restricted the technology.

      1. I am not a GM hater, however one has to be aware that GM, at least in Monsanto edition, is not all it has been trumped up to be. For one thing, Monsanto is charging higher prices for Roundup and Roundup ready seeds, thus skimming up for themselves large part of the assumed savings in agricultural production costs.

        Second, with regards to weed prevention, spraying crops with only one type of herbicide will quickly lead to weeds evolving to develop resistance to said herbicide. This indeed has already happened with Roundup and Roundup-resistant superweeds have developed. This means now farmers have to use and spend money on extra weed prevention on top of the premiums they are already paying for the Roundup ready crops and Roundup. In fact USDA estimates that introduction of those crops lead to an increase of herbicide use over the long term.

        A nice write-up on the subject from a respectable scientific institution can be read here

  6. You mention homeless people but even those in better situation (but not quite Silicon Valley wealth) get it pretty bad, even in seemingly dumber things. And it gets really annoying.

    So: I got some cheap tablet. Quite low-end, but OK it works. Its app partition is 1GB, the data partition is 4GB and has a slot for a SD card (currently has a 16GB one). Should be fine, right? Except because Android (at least the version it has) is completely unable to run any apps outside the app partition, and Google’s “critical” stuff needed to run Android is already eating like half of that, with a good bunch eating like 60~70MB each or so. Then there was the Facebook app eating up 117MB. Put together a handful more of apps (not many, maybe a handful?) and whoops can’t install any new app or even update the existing ones. Managed to get around that by uninstalling Facebook but gosh, why do so many of those Google apps eat up so much memory? And why can’t I make use of the rest of the internal memory that’s mostly underused? But I guess that won’t matter, Silicon Valley mentality is that you should keep getting the latest and bestest, who needs to care about those stuck with weaker stuff.

    And here’s where the actual problem comes. I complain about this (because I’m grumpy :P), and apparently the problem is not that the apps are bloated or that lots of memory goes to waste, but that I got an underpowered piece of s*** and that I should get a Nexus. But this is barely the most I could reasonably afford, better tablets easily cost the whole month worth of salary and are still low-end, the high-end hardware is outright not sold by stores locally and have to be imported directly, with all the headaches that involves. And then of course I keep getting comments about why don’t I just move to another country. Yeah, because I could simply just move, alright. I keep getting this comment like every other day or so, by the way.

    Apparently if I complain that I have a problem that could have been solved by the vendor just being somewhat more careful, it’s my fault for not having the same disposable income as a middle class person in a first world country.

    I just wish Silicon Valley wasn’t pretty much taking over all the Western world right now. Not only they completely ignore anybody who has different needs than them, they also change the perception of those who happen to be just well enough to afford the stuff they sell which indirectly hurts the rest of us in the long term. Kind of a big problem as the internet makes the world more globalized over time.

    1. Yup you are absolutely right. We really do not need any newer faster chips to do most of the stuff we want to do, if only coders would learn to write efficient code, but like you say, the assumption is that CPU and hard drive space are ‘free’ and nobody codes with any restrictions any more. its maddening.

Comments are currently closed.