How a cognitive bias is causing this tech bubble.

This post is not altogether different to my last one, on how we might be overestimating the capabilities of robots. The theme is the same. We are putting a frightening excess of faith in the future of technology.

We sit at an inflection point, extrapolating it to the stars.

Technological progress seems suddenly overwhelming. But there is reason to expect a breakdown in the recent rate of growth, reason to expect that we’ve grown deluded about the prospects of the silicon-based slice of progress we like to label “technology.”

Screen Shot 2018-03-27 at 1.04.17 PM

I wrote about this the other day in my regular column over at News.

Suddenly people are taking seriously all the following ideas:

Plans to colonise Mars;

Driverless cars taking over our cities within just a few years;

Flying cars;

Robots putting practically everybody out of work;

Artificial intelligence becoming so powerful it destroys us;

Cryogenics letting us come back from the dead;

Crypto-currencies taking over from money.

This is not just about people speculating on the future of a few companies. This is about believing the life of humans is about to change faster than ever before in human history. It is like a belief that we’re living through the agricultural revolution, the Renaissance and the Industrial revolution all at once — and all in fast forward.

Why so credulous?

Why do we suddenly believe technology will remake the future so utterly and swiftly? Partly because of a cognitive bias called the recency bias. We remember the recent past much better than the time before it. And in the recent past, technology has wreaked havoc on modern life. You’re reading this on a website that didn’t exist before 20 years ago.

In the past 20 years, the world has changed a lot. And technology has been a big part of it. But that doesn’t mean technology can change everything. The personal computing technology we all interact with daily has made it very obvious to us that technology can change very fast.

But this is a classic case of selection bias. If we try to measure the pace of technology by looking at the things that are changing very fast, we will get the wrong picture. We need to look elsewhere too.

If you tried to measure the pace of technology by looking at commercial aviation, say, what you’d discover is a lack of obvious progress. We used to have supersonic commercial aviation, but nowadays most of us fly around in Airbus A320s (a plane launched in the 1980s) and Boeing 737s (a plane first launched in the 1960s).

You can get a similarly glum feeling if you look at progress in fighting Alzheimers disease or Multiple Sclerosis. There hasn’t been any, despite a huge amount of effort. Likewise with the common cold — and we seem to be losing the battle against bacteria as they develop antibiotic resistance.

I don’t mean to say that technology won’t change. It can and surely will. Just to say that there is a certain wildness to the predictions of the future at the moment. People seem willing to believe just about anything, so long as it has a technology angle.

When the bubble finally pops, it will take with it not only the valuations of some of the biggest technology companies, but also a lot of utopian visions of the future.

In the News story I call it a recency bias but you might as easily call it an availability bias. We are very willing to believe technology can change the world utterly and quickly because in living memory personal computing has created very visible changes in our daily lives. (Maximally visible, but not necessarily maximally important – the famous hypothetical is whether you’d give up the internet before you gave up indoor plumbing.)


These cognitive biases have been allowed to grow unchallenged because of the peculiar financial circumstances of the times.

Some people argue the loose monetary policy of the last decade does not explain high asset prices, but I think they’re wrong. The simultaneous global bubbles in property, bonds and tech stocks almost certainly trace their roots to the low/zero/negative interest rates across much of the world, and quantitative easing that left developed economies awash with liquidity.

The money flood provided patient capital that gave companies with scant profits a long time to experiment and expand revenues. If you’ve ever taken an Uber using a 50 per cent discount, you’re using some venture capitalist’s money to improve your own lifestyle, while simultaneously propping up the impression that new tech is destined to remake the known world.

(For what it’s worth, Uber is pretty big improvement over taxis! But its major advantage comes from taking on a regulated market with colossal rents, rather than being inherent to the app.)

The money flood has propped up some far more dubious beliefs than the prospects of Uber. The faith certain investors have in Tesla’s ability to win a giant share of the “shared mobility market” (fleets of driverless taxis) is intriguing to me.

Path to price target

Valuing a junior company on the prospects of winning a large share of a market that doesn’t yet exist, using technology that is in its infancy? It seems, um … more optimistic than is prudent. If this kind of thing works for Elon Musk, perhaps he should also set up Red Real Estate and start selling rights to land on Mars.

The NASDAQ chart above explains why the cognitive bias we’ve developed has been allowed to progress so far. It’s a feedback loop from confidence, to investment, to expanding revenues, to stock prices, to headlines, to confidence.

And Bitcoin?!  … . Actually no. Let’s not even talk about Bitcoin.

(Non-financial evidence that technology really is changing the world, in the shape of temperature records and CO2 concentrations, doesn’t seem quite so influential on the mass mood. I leave it to the reader to ponder why.)

Eventually, the technology cycle of misplaced confidence and out-sized valuations will find it has reached the highest possible equilibrium and begin to tack backward.

Screen Shot 2017-12-01 at 11.20.16 AM
“The trend is your friend, til it bends.” – Anonymous.

It is likely to do that even absent a macroeconomic reason, but one is coming anyway.

Interest rates are rising in the United States and inflation is lifting. The anti-Keynesian Trump stimulus – adding fire to a booming economy – looks set to intensify those trends. The Fed is now slowly soaking back up loose money. This represents a clear and present danger to any asset whose value is not based on making real money right now.

If the market values of all those tech stocks fall, the stories they told about the future will suddenly appear thin. A pin will prick the bubble of credulity and the stories of inevitable autonomy, existential AI risk and imminent interplanetary expansion will fade from our front pages. The distance between the pssible and the probable will lengthen again.

So I’d like to place a stake in the ground and say we will look back on this era – with a TV show called Silicon Valley; a plan for Elon Musk to become the richest man in the world; non-stop headlines about drone delivery; and a relentless faith driverless cars were just a few months away – with a kind of nostalgia for a simpler and more optimistic time.


Published by


Thomas the Think Engine is the blog of a trained economist. It comes to you from Melbourne Australia.

13 thoughts on “How a cognitive bias is causing this tech bubble.”

  1. Interesting perspective; worth remembering that the industrial revolution was a period of 60-80 years, the agricultural revolution ~150 years and the Renaissance happened over three centuries! Given a long enough horizon, all those things are not only possible, but also probable.


  2. Drone delivery! This is something I am much more scared of than excited by. Ordering toothbrush refills and nasty sugar-free beverages in a free-market society is never going to fund efficient drone delivery systems. The only way a drone delivery system could be affordable is if one party controls the airspace and if this is the case I don’t think they’ll be delivering you convenient freedom and happiness to your domicile.

    Of your dot points of things people are taking seriously above, the artificial intelligence one is one that is interesting re timing. It might get not happen this year. Or in ten years. But when it does, what will happen? Self-imoroving AI development could be exponential. And how do you know the singularity has occured? The endless articles about lawyers being replaced by robots are just as annoying as drone delivery promises, because they seem to have a really lame understanding about the nature of change. If AI takes off properly the changes will be a lot more fundamental than out of work lawyers and I don’t think we’ll see it coming.

    Liked by 1 person

    1. > And how do you know the singularity has occurred?

      the plot to my unpublished sci fi manuscript (languishing deservedly in a long forgotten folder somewhere on this machine) is about how the singularity occurred, and the robots decided we were lovely people that we should be left alone; until we destroyed the environment so terribly that electricity was being rationed and they realised it was them or us.


  3. There’s been so much sudden talk about AI and robots in particular that I’ve been wondering if there’s some entity working overtime to *push* the fad into the media.

    (Oops – perhaps I shouldn’t use the word “entity”, as that makes it sound as if I think there’s an artificially intelligent robot phoning the papers. And maybe there is.)

    It must be some group, or groups, of people with a monetary interest, obviously. I was thinking of researchers who need to advertise their academic area to give them more chance of getting grants and continuing their career. A sad reality of life in this modern era where it seems increasingly hard to get funding from either the universities themselves or from the government.

    You see this with other areas of scientific endeavor, for example medicine, where you get the impression that researchers are on the phone to the press every time there’s even the slightest, most incremental advance in knowledge.

    But who knows, it could be developers wanting to get the world used to the idea of before there actually is some terrifying unleashing of unforeseen power.

    Liked by 1 person

    1. You are right there is financial motivation but it’s not people paying for things to be published. The push is actually a pull. What really motivates editors to publish things is when millions and millions of people click on them.

      That’s why we get celebrity gossip, house price, celebrity gossipy, weight loss, etc.. The pulling power of the clicks makes editors publish even more of it!


      1. That makes perfect sense, but I don’t know if that means it’s necessarily true. I’m not seeing those articles about AI and robots in the same place as the ones about celebrity gossip and weight loss. Because I don’t read the outlets with celebrity gossip and weight loss. Surely not all publications are entirely driven by clicks. Otherwise the BBC would be exactly the same as The Daily Mail.


  4. And, BTW, I didn’t suggest people were playing to have things published. I suggested that people in the industry were calling reporters with new developments because they needed things published to raise awareness of the topics and, thereby, to make them “sexier” and make it more likely that they would get funding for their research.


  5. James,

    Thanks for this analysis. It’s good to see. I don’t think I’ve read your stuff before – Newscorp is not an organisation I want to support any more but this piece could be published anywhere I suspect.

    Recency and availability must be almost the same functionally?

    Look I don’t agree on your explanation. Tech has surged ahead and it pervades life but it is an island of relatively proper progress in a sea of rusty conservatism. It’s it an example of market forces working somewhat properly. As it is, a lot of tech is now stalling. Ipads, smart phones and PCs have all climbed mountains of power and usefulness in recent years but are now crawling up gently sloping plateaus. Intel and nvida have slowed. Sales are slowing as people realise it is now mainly marketing driving upgrades, not massively better products. And there is nothing wrong with conserving stuff but it’s become too dominant. I think there is a widespread disappointment about the world and our progress in general. It’s not that certain things have recently created unreasonable over-excitement, it’s that there is a tide of concern and disappointment about the overall lack of progress. You mention aircraft and it’s a good example. But cars, houses, clothing, they’re all are the same. Incremental improvements over decades. Cost improvements and more digital screens but no basic improvements. Entrenched players in deeply ossified industries have doled out progress at the minimum possible rate.

    Prices have dropped for westerners due to international income disparities and advancements in sea trade. Great. But that’s not progress. It’s required the stripping out of economic diversity in each of those developed countries. It’s reduced options and possibilities for those people. The likes of Tesla have galvanized widespread support and/or piqued interest simply by disrupting that disappointing pattern by viewing one stalled sector through the ideas of one of the few relatively normally-progressing ones. I don’t even think batteries are the best option but I am very pleased Musk is having a crack.

    To be global and a little whimsical: a wave passed through in the 1960s into early 70s. We did some big things and by the late 1970s wealth inequality was it’s lowest. Since then wages have stalled and we’re all living in the white water. I’ve spent my whole life wondering about why we did so much and then retreated. Irrational exuberance is always a problem but interest in Mars etc is more about a yearning for a return to a previous era of thinking big. Of looking at frontiers and pushing into them. The basic issue is that it’s easier to stay home and consolidate. And certain people and certain kinds of thinking are better suited to safer pursuits. Let’s call them the people making the money now. The old eagles song said it, “There are no more new frontiers, we have got to make it here.” But making it here isn’t enough for humans. It ends up making us stale and increasingly concerned and impressed by increasingly smaller things. An old Ted Egan song said it as well, “Further out! We’re the overlanders, always further out! We’re undeterred by fire, flood and drought … as we take up each selection there’s a constant predilection to head in the right direction, further out!”

    Musk et al aren’t whipping up unreasonable optimism, they’re lightning rods for the anger and disappointment of people who can see we’re missing out due to the conservatism and lack of vision of the older generations.

    And this was a randomly long reply I know :-). Your article tapped into something I think about a bit. Cheers. John


  6. Hiya, I just subscribed to your blog today because I read an article by you in the SMH and followed the link. I am only recently interested in economics and mainly only in how it effects the actions of the Australian government on Climate Change.
    But my husband is in IT and is a keen follower of Elon Musk – both his roles in Tesla and in SpaceX. My husband is no dolt when it comes to knowledge about rockets, electric cars, AI, etc. While it’s not his line of work, it is has been an ongoing hobby for the last 40 odd years. And from what he watches, I’m not sure that Tesla’s potential should be so swiftly doubted. I know a lot of economists struggle with Musk’s business model and possibly even his vision. He is a disruptor and that makes him hard to predict. But from what I hear from my husband (and his equally interested brother), the technology is pretty incredibly sound. Musk may not always be the smartest guy when it comes to PR, but he’s definitely a thinker. And he seems to think problems through from the beginning. Unlike many other companies, Tesla hasn’t accepted that Waymo is the way forward with self-driving cars and has chosen to pursue his own path. Not only that, but there is a whole fleet of thousands of cars out there sending Tesla data all the time. Waymo’s approach was “the thing” 10 years ago but it seems to have reached a dead end.
    All I’m saying is, economics is about modeling and predicting but, in order to predict, it’s handy to know something of the product/service you’re forecasting for. Tesla is about brand new technology. It would be interesting to look into history and see what economists said about Ford’s first factory or about telephones when they first came out. My husband remembers doubting that LCD TVs would ever surpass CRTs because he thought they would never be affordable. Turns out he read the wrong crystal ball and he learned something from that. :-)


    1. Hi

      I agree Musk is hard to predict and that anyone who predicts his doom is taking a risk.

      My view is this: the recent increase in loose capital in the global economy allows companies to take more risks. Risky companies are being backed by more money than in the past. This explains why a lot of tech companies have got a lot of capital, and why a lot of companies have very high price-to-earnings ratios.

      This extra capital chasing risk has two effects. One is that more of the capital will be lost. It will be sunk into MySpaces and Pets.coms and Theranoses. The second is that the extra capital actually reduces the risk those companies face. It is patient capital and it allows them long runways so they can “fail” for many years as operating businesses before the failure shows up in their stiock. The big question is how much does it reduce the risks? i.e. For how many decades can a company like Tesla lose money? We’re at 1.5 decades now. Can they do another decade?

      A couple of things I’d like to take you up on:


      >Musk may not always be the smartest guy when it comes to PR, but he’s definitely a thinker.

      I’d say Musk’s great strength is PR. Have you ever seen another company get so much attention? Or raise so much in deposits? He has a way of selling everything. One that I always return to is Chill Mode. Any other company would call it Eco Mode. But he’s a genius at this sort of thing. He can even make driving slowly to conserve range seem fun and exciting.


      > It would be interesting to look into history and see what economists said about Ford’s first factory or about telephones when they first came out. My husband remembers doubting that LCD TVs would ever surpass CRTs because he thought they would never be affordable. Turns out he read the wrong crystal ball and he learned something from that

      I agree it’s important to not doubt every technology. But you can easily over-correct and start to believe in every technology. We still have to apply critical thinking.

      I think battery vehicles will be a good business if they can get the costs down just a little more. I think autonomous driving will be an INCREDIBLE business once it is solved. The biggest money-maker imaginable. But I look at Tesla and I think they took on so much debt they will face a terrible reckoning before battery costs are really low, and long before autonomous driving is safe in most situations.


      1. Hi. Thanks for your quick response.

        I don’t really subscribe to the theory that any attention is good attention. Nearly every day there are negative stories about Tesla. The good ones are harder to find but they’re there if you look. And if you look at everything (like my husband does), the positive ones seem more convincing than the negative ones.

        But what I meant when I talked about Musk’s PR was his twitter feed – his optimistic time frames, his terrible “paedo” comment about the Thai cave diver, his misplaced claims about having “found funding” to go private, etc. To me they reflect poorly and they weren’t necessary.

        I can see your point about the “patient capital”. Companies like Uber and Amazon have benefitted from that. But I think it’s worth noting that Musk is also “the brains” behind SpaceX. SpaceX doesn’t get nearly as much negative press as Tesla does and my guess is that that’s because they’re not disrupting their industry. And in them, we can see the success of Musk’s technological prowess. He has proved, in the life of SpaceX, that you can re-use rockets. He has earned the trust of NASA and he has taken back into the USA, the job that has (since the shuttle retired) been outsourced to Russia. SpaceX has not been floated on the public market so the opportunities for damage are fewer but I think, even if they were, they would have fewer enemies because they do not pose a threat to the oil industry.

        As you say, Tesla has had the patience of its supporters for 15 years. But is that really so long for a car company to get up and running? I don’t know. I don’t remember what it was like when Audi or Hyundai began. They weren’t notable because they weren’t doing anything that different. And it’s worth mentioning that none of the “established” car companies are making any profit on their electric cars while Tesla is making a 20% margin on every car they sell.

        Anyway, I’m sorry. Your blog was not really about Tesla. But I am interested in your ideas and I look forward to reading more blogs. I have a friend in the Finance Industry and she told me about a podcast she listens to which has a pro-Tesla economist and an anti-Tesla economist and she can’t pick which one she believes more. So it’s clearly a divisive area.

        Thank you for your patience.

        Cheers, Jo


  7. Taxi’s used to work very well for me. I’d walk out of my home or office, flag down a passing cab, and be on my way. Flagging down a cab used to be a standard operation. And if a busy cab went past and saw you, they’d call in the job. Then it was replaced with booking agencies. You’d call for a cab. Eventually, cabbies would no longer call in jobs, because the booking agencies wouldn’t take them — they got paid for phone bookings. And drivers stopped looking for jobs — they got jobs from phone bookings. And drivers stopped roaming for jobs — they’d sit and wait for a phone booking.

    And taxi licenses were limited. The service is driven down, the price is driven up. Drivers pay half their income to rent the license – they can’t afford to be roaming the streets looking for work.

    Uber works better than taxis /because the taxi system stopped working/.


Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s