No. 93: What if Uber disappeared? 🗑🚗

A story in two underwhelming acts

Act I

As we departed the Royal Albert Hall last weekend, heavenly Bach still soothing our ears, we were blindsided by a pixellated sight no milliennial should bear.

We couldn’t get an Uber.

Or even a Lyft.

Bolt? Fat chance.

After a notable period of sighing and cursing (3-4 minutes, tops), I resigned myself to more traditional modes of transport. One foot in front of the other, scenery slowly passing by, the wind barely bothering my pristine locks, I walked.

Then we took a train.

An underground train - and that part is important, because I might as well have descended to the underworld for a tête-à-tête with Hades himself, such was the white-hot atmosphere down there. Lucifer’s embrace would offer chilly relief, were it to follow a journey in that infernal carriage.

“Tonight, no Uber. Tonight, I am a Tuber”, I thought, the heat only slightly melting my perspicacity.

Act II

I was back in Belfast last week to see family, when I was blindsided by a pixellated sight no milliennial should bear.

I couldn’t get an Uber.

Etc. and so on.

I took a bus. I got to where I was going.

I also got to reconnect with some of my fellow Belfastians, but apart from that it was fine.


Uber is having a bit of trouble at the moment. Uber always is, but this is a little different.

In London, a number of people have reported very similar stories to my own. In more prosaic terms, admittedly.

Apparently, new services are offering bonuses to entice drivers away from Uber. In the UK, a lot of drivers have also left the country due to the self-imposed sanctions of Brexit. And lo, we walk, we Tube, or we wait. It’s hell, I tell you.

Except it’s not altogether so terrible.

One very quickly finds an alternative and, while our brains have quickly adjusted to the soothing convenience of modern living, the tiniest bit of friction can still be endured. After 3-4 minutes of sighing and cursing, of course.

For Uber, the fault lines run deeper and wider than just London and Belfast, however.

Their business model accepts huge losses to subsidise those affordable rides we all take for granted.

So it seems timely to ask:

For how long would we care if Uber just disappeared?

For all the “disruption” talk, what would be its lasting impact on our cities?

There was an article on Medium recently, called “End of the Line for Uber”. It was written in that overconfident, definitive style that must get clicks on Medium. Beneath the hyperbole, there was some substance to the argument.

The author argues that we would see a concrete impact in the event of Uber’s demise, albeit not a positive one. Uber has restricted investment in public transport infrastructure and increased congestion, he points out, without paying drivers sufficiently or protecting passengers at all.

“I live in Burbank, where Ubers were never more than 5 minutes away. Now, a 30 minute wait is common — and the fare is comparable to the licensed taxi company, which is literally one app away from achieving feature parity with Uber.”

This last point is particularly apposite.

I was amazed last week that local taxi firms in Belfast now have Uber-style features on their apps. I was flabbergasted that they even had apps.

I also recall reading that in Paris, the licensed taxi drivers had taken etiquette lessons to help compete with Uber. If you’ve been in an Uber, you’ll know that that’s a low bar to limbo under, but the surly Parisians managed it.

They also invested in an app that would offer a booking service with a five-minute time window. Uber typically offers a ten or fifteen minute reservation window.

That’s the thing with all this bro-friendly “disruption” talk. The process is ongoing; the established businesses can learn and adapt. If the self-appointed disruptors end up raising prices just to try and break even, their business model was unsustainable all along. They also lose a very appealing, competitive advantage when they do so.

I wouldn’t go so far as that Medium article’s brutal conclusion: “The sooner Uber dies, the better.”

But I would say that if it ever did go the way of the dodo, its residual impact would be seen most in its rivals. The customers would get along just fine. Our cities might even improve, too.

Chart of the week 📊

I took a look at the ecommerce revenues of the biggest retailers in the US, in 2019 and 2020, then made some charts.

You will expect Amazon’s continued dominance, although the scale of that dominance is striking:

And here again, with figures and % increase from 2019 to 2020:

Many of these retailers shifted business from physical stores to their websites in 2020, of course.

Amazon has a significant advantage here, but you might also say it is in a position to be shot at.

That’s why I find it so intriguing that Amazon is starting to license its fulfilment services to other companies. It knows that they will start to get their act together. They may even work together to create alternatives that can compete.

To stave off the threat, Amazon would like to offer them access to its superior software and logistics. In return, it will receive a lot of valuable data on top of the additional revenues.

Amazon’s rivals will surely consider that trade-off carefully.

On a smaller scale, this story relays a similar situation with a retailer here in the UK:

  • Whatever Next? How a middle-of-the-road high-street chain became a retail powerhouse - The Guardian

Copy, Waste 🗑

TikTok is building an Augmented Reality effects studio. Yeah, like Snapchat and the other ones.

Not much else to say on that uninspiring story.

The Data Hater

It’s back! A semi-regular feature in which I rag on sub-par data viz.

This time: the neverending pie chart.

Triggered emails? You sure triggered this emailer, that’s for sure.

Other Reads

  • You are a network - Aeon

This is superb. It’s all about the “emerging theory of selfhood” that extends beyond the traditional ‘container’ of the individual body. Thought-provokin’ stuff.

  • The architecture of the 2020 Tokyo Olympics - Dezeen

  • CodeVox turns spoken, natural language into lines of code - VICE

No. 92: 🏀🪙

The blockchain-based fan token platform you never knew you needed


This week we’re looking at Socios fan tokens. Football clubs, basketball “franchises”, and Formula 1 teams have signed up to this blockchain-based system and it has been all over the news recently.

Arsenal, AC Milan, and Barcelona are all on board.

If you recall the mercifully short-lived ‘European Super League’, you’ll know that when this lot get their heads together, sensational ideas tend to follow.

(Oh and speaking of F1, I snuck out hi, tech. 91 earlier. After the… lukewarm reception for my football specials, I thought it best to avoid emailing a motorsports special so soon. I can only lose so many subscribers. lol.

You can read the F1 special here if you like, though.)

I know, I know, it’s another sports edition of hi, tech. But it’s really about the tokens and the blockchain, rather than the teams. I have just read that last sentence back and I do not know if it makes this any more appetising a read.

As a primer, you could always check out the hi, tech. special on NFT art from a little while ago.

What is

Socios (Spanish for ‘partners’) runs on the Chiliz sports blockchain and it creates fan tokens for a range of organisations. Fans can then buy the tokens, for dubious reasons I’ll explore below.

Fan tokens have generated over $200 million in sales worldwide already and Socios is moving aggressively to seize the first-mover advantage. Their logo will ruin shirts from Valencia to Milan this season, with many more unlucky clubs sure to follow. Argentina’s top football league will officially be called Torneo for the 2021/2022 season. 

Socios is not the only company aiming to capitalise on the cryptocurrency cash-grab. Sorare, for example, offers blockchain-based fantasy football and it has become insanely popular already.

Instead of each sports club coming up with their own blockchain-based fan token system (because obviously, they need one), Socios offers a platform that can host all clubs and teams. Socios then creates and sells team-specific digital collectibles, a little like the NFT art and NBA Top Shots we looked at in hi, tech. 79.

The all-important network effects

Socios’ CEO gave a transparent take on his company’s growth strategy on Twitter this week:

Has this guy taken one of my business school courses?

  • Network effects ✅

  • Platform ✅

  • Ecosystem ✅

All in one tweet, too. +100 Disruption Points for you, Mr Dreyfus.

It is a tried-and-tested strategy, plus the emotional heft of these clubs will attract more and more fans over time. What they really want is to get to the point where fans whose teams are not on Socios start asking for their clubs to sign up.

The very name ‘Socios’ is intended to create the sense that fans are in on this with the company. If Socios wins, everyone wins.

Except, they won’t - and you know they won’t.

There are echoes of the Football Index farrago here.

As this Guardian report put it,

“By blurring the lines between gambling and investment, the suspended company Football Index has left many in debt.”

Football Index was pitched as a way to make money from your football knowledge, by buying shares in promising players and selling them at a profit once the players gained in value. They sponsored some notable teams and ran expensive ad campaigns on TV, radio, and the London Underground.

But the shares were not tied to anything tangible in the ‘real’ world of football. They were driven by supply and demand within the Football Index app. It didn’t take much for it to fall apart.

With all of these top teams and leagues involved, one would assume there is some substance to Socios.

And that’s precisely what they want you to think: Socios is betting that if it acts like a real thing and it gets some big names on board, it will gain the legitimacy it needs to take off. It has set aside $50 million to add more US sports teams this year, too.

How do the sports teams make money?

This is still opaque, although Socios recently announced that the platform had generated $150M in revenues for sports teams in 2021.

There are some statistics about the mist popular fan tokens on there, too:

Each club has an initial, one-day period known as a Fan Token Offering (FTO) when the tokens are sold at a fixed price. After the FTO, they are listed and subjected to good ol’ market forces.

The screenshot above was taken just after the Galatasary FTO, hence their exalted status in the list.

Some of the price increases are highly significant, which raises the question: Is this a cryptocurrency exchange for speculators, or an actual fan platform? Can it be both?

Let’s take a look at how the app works.

The experience: How does it work and is it any good?

First of all, they ask for your money. They’ll take your real currency and turn it into Chiliz:

The app says,

“Think of $CHZ as like a foreign currency when you go on holiday”.

I’m sure this is intended to reassure, but it just makes me think of that time Homer traded his real money for Itchy and Scratchy dollars, then couldn’t even spend them in Itchy and Scratchy Land:

And when they put it like this, I can’t think of any other way I would wish to spend my debit card balance:

Good of them to take credit card, too.

There isn’t much going on in the app, in all honesty. You can look at different teams:

To see their prices, trade coins, or join votes. More on the latter very soon.

Modern sports fans are forever chuntering on about how “big” their club is on social media, so they will be competitive about the price of their tokens. Essentially, all of their jumped-up hyperbole will fuel the Socios token market and it seems like an inexhaustible resource.

Ok, I got my Chiliz. I bought my tokens. How can I use them?

Having retained some semblance of my sanity, I didn’t buy the tokens to test this out.
But according to the Socios app, you gain:

  • Unrivalled access to your team (As yet, undefined)

  • Once in a lifetime experiences (Being taken for a mug isn’t a once in a lifetime experience, but whatever)

  • Quizzes and games (Woohoo!)

  • The ability to vote on “official club matters”

Now this last one sounds intriguing, but what does it entail?

It got me thinking of this quote from Arsène Wenger:

"In the next five years, it might happen that social media substitutes players during a match.

They will have a hook-up at half-time and determine which players get substituted and who will be brought on during the second half.”

Is the future here?

Well, I had a look and this was the top poll:

So you’re telling me I can turn my money into Chiliz, turn my Chiliz into tokens, then vote on two very similar bus patterns for a mid-table La Liga team? I’d rather have Itchy and Scratchy dollars, thank you.

You can trade for another team’s tokens, if you have a specific interest in the livery of their autobuses:

Or if you want to try and make some money. Speaking of which…

Man: A symbolic animal

I am increasingly won over by the argument that man is not so much a political animal, as Aristotle had it; but rather a symbolic animal, as Ernst Cassirer proposed.

Sports clubs trade on this tribal symbolism.

Socios is a straightforward way to show your commitment and they don’t half ham it up.

The current hashtag they are pushing is:


And what it lacks in legibility, it more than makes up for in volume.

You could buy actual shares in some of these clubs - Juventus and Manchester United are publicly listed. But these tokens aren’t about that kind of ownership. They sit at that trendy nexus between crypto, social media, and gaming.

In all the talk of the European Super League, one phrase stuck with me. The clubs referred to their loyal, local supporters as “legacy fans”.

There is something quaint and outdated in the connotations of that phrase, isn’t there? The same fans who buy season tickets, travel to away games, and support the club through thick and thin are the “legacy”. The future belongs to a different kind of fan.

Socios fits rather neatly with this vision.

Many people struggle to conceptualise the relationship between cryptocurrency and the world we already know.

This form of cryptocurrency at least has some symbolic meaning. I am #MORETHANAFAN because I have bought the club’s tokens and I can vote on club-related matters. You would imagine the clubs and F1 teams will offer more access than voting on bus colours in future, too.

As some suit at Arsenal put it:

“We are committed to finding new and innovative ways for all our supporters - whether local or international - to get closer to the club."

- Arsenal commercial director Peter Silverstone.

It is a get-rich-quick scheme dressed up as a global partnership, but he can’t say that.

Sure, clubs could manipulate fans into spending money on tokens in return for a vote on which urinal cakes to buy for the stadium toilets.

Or, they could engage directly with the existing fan groups that would love to have input on meaningful matters. They could offer fans real ownership of the club, as some teams (Rangers, for example) have started to explore. West Ham fans protested the club’s Socios link-up and the club listened - they decided against signing up to the platform.

It is perhaps overly romantic to expect anything different from the clubs that signed up to the European Super League. It is also undeniable that sports fandom is changing rapidly and the clubs need to change too.

I recall watching a documentary about Manchester City and the club’s chief executive was shocked at how little revenue the club gains from each fan worldwide. Global fans want a way to engage with the clubs online and the clubs would be foolish not to monetise this desire. The stadiums can only hold so many fans, after all.

But you do have to fear that if the whole thing comes crashing down, the well-intentioned fans will be left holding these tokens while the clubs move onto the next scheme.


Tech Bites

How video game artificial intelligence is evolving - The Guardian

Can Pittsburgh Make ‘Mobility as a Service’ Succeed? - Bloomberg

🏎 No. 91: Williams Racing - Is The F1 Team Getting Back on Track?

An Exploratory Data Analysis

A small group of teams has an outsized impact on the collective imagination of racing fans. Certain images spring to mind when we think of drivers, circuits, even decades. The supremacy of the Williams car in the 90s, Rothmans livery and Renault engine, falls into this category.

The immediacy of memories inevitably fades. If the stock is not replenished, exhilaration gives way to sepia-tinged nostalgia.

Williams is one of a group of five teams that won every Constructors' Championship between 1979 and 2008 and every Drivers' Championship from 1984 to 2008. Its place in the sport’s heritage is unquestioned; the team’s glory days have shaped the contours of Formula 1 as we know it.

The team’s recent travails make for a stark juxtaposition, posed alongside these heady achievements. Williams’ last win was in 2012. The win before that was in 2004.

Google offers these merciless suggestions in response to a query for ‘Williams Racing’:

The team founded by Sir Frank Williams in 1977 is now in the hands of Dorliton Capital. The recent takeover provides a needed change of strategic direction, leading many to ask whether Williams is back on track.

I took a look at their storied past, both distant and near, to see if the data suggest an overdue return to the leading pack.

Williams Racing: The Past

Williams has won a total of 9 constructors’ championships, second only to Ferrari.

Williams is clearly part of an élite club here, alongside the likes of Ferrari and McLaren - but the former retains a healthy lead.

Ferrari is the only team to have run a car in every championship since F1’s inauguration in 1950, although they did miss the first race due to a dispute with organisers.

Williams has run a car in 754 races since 1977 and, strikingly, they have finished races in 1st position more often than any other position. Williams has 114 race wins, amounting to a 15% race win ratio.

Add in 128 pole positions and the pedigree of the team becomes clearer still.

The chart below shows the cumulative number of championships won by each team over the years, including both drivers’ and constructors’ titles.

Ferrari extends its lead further, by this count.

Ferrari won multiples driver’s championships in the 1950s, but did not win the constructors’ title in those years due to its lack of existence. The constructors’ championship only came along in 1958.

From this chart, we can see a simplified story of F1’s past. Williams had a very productive period in the 1990s, securing five constructors' titles in the space of six seasons with an engine supplied by Renault.

After 2013, we can see the impressive rise of a Mercedes team that has dominated the v6 turbo era. Between them, Red Bull and Mercedes have won every constructors’ title since 2010. In 2009, Brawn won the title, and then Mercedes bought them to kickstart their F1 entrance.

We can also note the fallow periods that all teams suffer. Ferrari had barren sequences from 1961-1974, 1979-2000, and 2007-2020. Let’s be honest, they’ll soon add 2021 to that list.

Williams is undoubtedly enduring one such period right now. The chart below shows the frequency of positions the team has held in the constructors’ championship this century.

Third is the most common position (4 occasions), with the last third-placed finish coming in 2015:

This tells only a small part of the story. Two teams (Mercedes and Red Bull) have led the field by such a margin that third place may not mean the team in third is within striking distance.

The chart below shows the percentage point deficit from the leading team each year to Williams:

In 2015, Williams finished third on 257 points. Mercedes won the championship with 703 points. Williams amassed 36.5% of Mercedes’ points, creating a 63.5% deficit from the leader to Williams.

In 2003, Williams finished second on 142 points, close behind Ferrari on 158 points. It is very important to state that the points awarded for a win are markedly different today (it was only 10 points for a win back then), but the percentage points deficit calculations do still paint a representative picture.

We can look at how Williams has fared against the leading team each season to gain a different view of the team’s stalled progress:

These graphs show the accumulation of points across each season. Mercedes gathers points at a reliable, rapid rate, while Williams has ground to a halt of late.

As highlighted in some excellent work here by James Trotman, the pattern is playing out again this season:

Trotman displays some illuminating graphs from the individual races, too. For example, we can see here how the leading pack breaks away from the rest and extends its lead throughout the 2020 French Grand Prix:

You can only really understand F1 with your financial hat on, of course.

In recent seasons, Williams has had one of the smallest budgets on the grid and this correlates closely with its performance at the chequered flag.

What happened to Williams? That happened.

The new annual budget cap of $145 million has incentivised smaller teams to stay in the sport. Admittedly, some items are exempt from that cap, including marketing, driver salaries, and about 20 other line items.

Williams: Signs of Progress

With sweeping changes on the horizon in F1 (more on that below), Williams will be keen to show signs of progress this season.

So far, this progress has been manifest in qualifying - particularly from George Russell:

In fact, Russell is the only driver to have beaten his teammate (the Canadian, Nicholas Latifi) in qualifying at every race so far this year.

The challenges mount for Williams when the lights go out, however. Russell loses an average of 1.44 positions on the first lap alone, perhaps suggesting he is getting everything he can out of the car for one excellent qualifying lap.

Russell has lost a total of 13 positions on the first laps of races this year so far.

Still, Williams did post the quickest pit stops at the British Grand Prix, so they made up a little bit of time there.

Below, I have plotted the total number of laps four drivers (Hamilton, Verstappen, Russell, Latifi) have spent in each position this season:

It may seem cruel to make this comparison, but I was struck by Verstappen’s command of this year’s races. He has led for 70% of the total laps this season, and would surely have added to that total at the British Grand Prix were it not for that first lap incident with Hamilton.

Further down the order, we see that Russell routinely outperforms his teammate, Latifi. Latifi, 26, is rumoured to bring in sponsorship money of $20-40 million a year. His father is the founder of the third-largest food company in Canada, Sofina. He’s not a bad driver, but Russell is clearly the superior talent.

Should Russell continue his recent form, he will only attract increased attention from the Mercedes team.

Williams has forged closer ties with its engine supplier (Mercedes) and the rumours that Russell will take the number two seat at Hamilton’s team continue to intensify.

Yet Russell’s form could be a win-win for Williams. Their number one driver may leave, but he will show what the car is capable of in the process. Williams has ambitious plans for next year and these will be fuelled by improved performance for the rest of this season. If Russell takes the Mercedes drive, that would logically leave Valtteri Bottas (a former Williams driver) open to offers.

The 2022 Season

The 2022 season will see a host of changes to F1 cars, in reforms that many see as the largest in the sport’s history. As we saw above, there is a very small group of supremely well-funded teams that dominates F1 today. Once they get to the front, there is limited overtaking.

That limits the appeal of the sport for a wider audience, both in terms of viewers and sponsors.

The budget cap will force teams to think creatively, but 2022 also brings new technical rules to try and improve the spectacle.

As Pat Symonds, the Chief Technical Officer at Formula 1, put it recently:

“'The wake' is the crux of the overtaking issue. When running at speed, F1 cars create turbulent air, which is dispersed outwards so as not to disturb the rear wing. The resulting wake affects the following car by significantly reducing its downforce (which ‘sticks’ cars to the road) and causing vital systems to overheat. It is therefore very difficult to get close enough to attempt an overtaking manoeuvre or even follow another car through corners.”

Formula 1 has hired a lead designer to help with the car’s aesthetic appeal, too. The F1 website has a great write-up of the new car here.

We began by looking at the history of F1 racing, which showed that major upheavals in the rules can shake up the order. When new rules come into play, the teams battle to take advantage wherever possible. Brawn managed this in 2009 and Mercedes then seized the initiative in the new era from 2014 onwards.

If we look at that 2014 championship, we can see the performance of each team versus their predicted performance from the previous year’s championship. Anyone above the line outperformed expectations:

That is not to say that Mercedes will repeat the feat this time; rather that it is likely one manufacturer will crack the code earlier than the others.

Williams may not be at the front of the grid any more, but it can take advantage of the improved slipstream behind Mercedes’ cars in 2022.

If it can continue to post positive results, not only in qualifying but also in races, 2022 can be the true beginning of the Williams renaissance.

No. 90: 🤖 Github Copilot

+ 🐙🔮 Clark the Octopus, TikTok soccer streams, NY Phil data

🐙 Remember Paul the Octopus? 🔮

He was the soothsaying cephalopod who foresaw the results of the 2010 World Cup. They’d give him two boxes of food, each with a flag on it, and whichever one he ate from was his “prediction”. He kept getting it right, right up to the final.

Google made this Doodle of him for the 2014 World Cup:

Paul lived as fine a life as any octopus could wish for, rivalled in luxury only by Squiddly Diddly.

The Spanish state offered Paul protection, his zoo in Germany made a statue of him, and a group of businessmen in Galicia offered a €30,000 transfer fee to make him the headline act at the Fiesta del Pulpo. You have to wonder how many live octopuses feature at this fiesta, don’t you? But people made exceptions for Paul; he earned eternal respect through his prophecies.

Sure, some said Paul was just lucky. And yeah, the President of Iran called him “a symbol of Western decadence and decay”, but the avant-garde will always threaten the status quo.

🤨 We revere those who can see the future. They hold a special place in the collective imagination. They are justifiably rewarded with enough riches to make Croesus blush.

Where I’m going with this is, this octopus got what was coming to him. TV appearances, all the clams you could ever eat, immortality in cartoon form.

And with that, let us recall Friday’s special edition of hi, tech. and a prediction even Paul would approve:

Heck, we even picked out Italy as the team to back before a ball was kicked.

If Google would like some inspiration for my upcoming Doodle, I offer this picture taken in my home last night:

Hope you all enjoyed a highly entertaining Euro 2020 final, and that you made some serious bones (money) from the Italy predictions.

Commiserations to England. They’ll have a strong chance of winning next year’s World Cup. The racist abuse of the players who missed the penalties is an absolute disgrace, too.

🔮 One more prediction: I’ll give the predictions a rest for a while. The chances of getting this lucky again are so slim, even Paul would steer clear.

Share hi, tech.

👨‍✈️ Github Copilot: What You Need to Know

Now, let’s take a look at Github Copilot.

I chatted with Chip Huyen, lecturer in machine learning at Stanford University, to get the inside track. Chip produces a lot of fantastic content and is very much worth following on LinkedIn here, and/or Twitter here. Her Machine Learning Interviews Book is superb, too.

What is Github Copilot?

Github (owned by Microsoft) and OpenAI have partnered to create Copilot, an AI-powered pair programmer. Microsoft invested $1 billion in OpenAI last year, too.

Github Copilot is powered by OpenAI’s Codex, which translates natural language into code.

Copilot is embedded in the Visual Studio Code editor (in beta for now), and it makes automated suggestions to programmers as they type. In the example below, the user provides a natural language prompt for context and begins typing code, before Copilot steps in to provide recommendations.

So this gets us a little closer to what this product really does. Think of the autocomplete suggestions you see in email software and you’re on the right track. Except that this is a much more complex problem to solve, with deeper implications if the solutions are flawed. Copilot also aims to go further than autocomplete, by inferring context from code to suggest altogether new functions.

It looks like this:

The idea of AI-assisted code is not new (hey, is any idea completely new?), but with Microsoft behind it Copilot has caused quite a stir. There are over 1250 comments on this thread about Copilot on Hacker News, for example.

Copilot: The experience

There is a waiting list to trial it and you better believe this joker is nowhere near the top of that list. I’d pencil me in for a tenative start date of 2025.

So I asked a real expert, Chip Huyen, about the experience of using Copilot. She says there are potential uses for Copilot beyond the obvious ones, for example with help writing documents. Codex’s natural language technology could prove its worth here.

Huyen also notes that the UI has room for improvement in its current form. “It’s very distracting to have code suggestions jump out at you when you write code”, she adds.

From what I have seen of the product in the wild, programmers can provide a short prompt and then Copilot provides suggested blocks of code. The programmer can scroll through these to find one that works. It’s easy to see how this could be distracting.

How it works

Copilot generates blocks of code, some of which have never been seen before. How does it do this?

Well, it is trained on millions of Github repositories, along with other open-source data.

In the FAQs on the Copilot site, they do attempt to address this directly:

They say that about 0.1% of the time the code will be taken verbatim from the training set. The other 99.9% of the time, it is using that training set as the basis either to synthesise different pieces of code or generate something new.

Some programmers would say the boundary between those two camps is not as fixed as Github would like you to believe. Ultimately, it is dependent on a training data set pulled from work by millions of real-life, human professionals.

I asked Chip Huyen about this and she said,

“It doesn't seem like the training process was too sophisticated and it looks like the model just overfits to training data (spitting out exact code from the training data), so there's risk of injecting bad/malware code into your product.”

Which brings us to our final question: Will Copilot replace human programmers?

And the answer seems like a hard no. It will certainly improve, both in its UI and its effectiveness, but this is intended as an assistant and it relies on a programmer’s expertise. Their job is about a whole lot more than typing out repetitive blocks of code, too.

The intellectual property questions are intriguing and Microsoft is adamant that the training data falls under the “fair use” category. There is precedent here; for example, when Google Books successfully argued a similar case.

For now, that’s what you need to know about Github Copilot!


📺 Social Streaming

When’s the last time you went on the Burnley Football Club website?

Actually: When’s the first time you went on the Burnley Football Club website?

For me, the two answers are the same: Yesterday.

But went there I did, all in the name of this announcement:

“Burnley FC Women will become the first team to have their games streamed live on TikTok – which will also become the team’s sleeve sponsor – as it undergoes significant expansion as part of the club’s new women’s football strategy.”

Amazon already streams sports and there have been rumours about Facebook buying football rights, but this is still a striking moment. For one, it’s a women’s team that is taking the lead by showing games on TikTok.

Also, it is a sign of what is to come on TikTok - don’t count them out of the so-called “streaming wars”.

📈 Exploratory Data Analysis of the week

I found a data set with every performance by the New York Philharmonic and I do miss going to see those guys, so I took a look.

The New York Philharmonic played its first concert on December 7, 1842. There’s a lot of data to look at, is what I’m saying.

For a first glance, I pulled out the most popular composers from 1842-present and not so surprisingly, Ludwig van takes the crown.

I’ve started looking at how they have diversified over the years too, in terms of composers but also venues.

Share hi, tech.

😔 The Data Hater

From the sublime to the ridiculous. We chew all the meats of our data-driven stew here at hi, tech.

I love Penguin and they make a fantastic point here - not that the visualisation would help you realise it:

🤖 Tech Bites

This week, I’m sharing some articles about the state of the online world.

  • Social media is broken: Here’s how we can fix it - MIT

“Social media is rewiring the central nervous system of humanity in real time,” said MIT Sloan professor Sinan Aral, who led the event. “We’re now at a crossroads between its promise and its peril.”

“The glue that holds humanity’s knowledge together is coming undone.”

  • The AI Wolf That Preferred Suicide Over Eating Sheep - OneZero

“An unemotional response from the AI of a simple game sparked off an emotional response among Chinese netizens, causing the kamikaze wolf anecdote to spread throughout various local social media platforms.”

Lots of excellent research in this guide from Conductor, based on international search data.


☝️ And finally…

It’s back! The Pudding’s AI New Yorker caption competition - Pudding

My dream is to win the New Yorker cartoon caption competition and this AI is aiming for the same.

This is likely to be of greater significance to me than Github Copilot, if we’re being honest.

See you next time!

Euro 2020 Final: A Data-Driven Preview

It's Coming, Rome 🤌 🇮🇹

Welcome to a very special edition of hi, tech.

I’ll warn you now: It’s heavy on the stats. The word “if” also does a lot of heavy lifting.

Oh sure, I’ve peppered in the odd jokette here, a tired stereotype there, but it’s mainly a hard-hitting soccer-based analysis.

More traditional fare is on the way this Sunday, when I’ll be reviewing Github Copilot.

Now if you recall our Euro 2020 tournament preview edition, we suggested Italy as potential winners and I do hope you placed that tidy bet. We made no other predictions in that edition. Nope, none at all whatsoever and let’s just leave it at that. 😉

The final is on Sunday and we have 🇮🇹 Italy up against 🏴󠁧󠁢󠁥󠁮󠁧󠁿 England. I live in England and I can assure you, it is hard to ignore that fact. I have lived in England for many years, but I remain as English as a taco.

The analysis from high-paid pundits in this country can be distilled down to:

“We’ll do ‘em, easy - ‘cause we’re England, innit”


And I’m not even exaggerating much there.

So I’m hoping there’s room for a more reasoned take. This will be the fair, balanced assessment you’ve come to expect round these parts. 🤨

📈 The data below is taken from:

  • UEFA

  • Statsbomb

  • The Analyst

  • WhoScored

If you enjoy this edition, please do take a short second to forward/share:


This remains a very free newsletter and some research goes into it, believe it or not. 😲


OK, let’s-a-go!

Predicted teams

🏴󠁧󠁢󠁥󠁮󠁧󠁿 England

(with WhoScored ratings out of 10 for tournament performance so far):

👶 Average age: 25

🇮🇹 Italy

💀 Average age: 28

Those ratings out of 10 are calculated like this, by the way:

There are over 200 raw statistics included in the calculation of a player's/team’s rating, weighted according to their influence within the game. Every event of importance is taken into account, with a positive or negative effect on ratings weighted in relation to its area on the pitch and its outcome.

An example: An attempted dribble (event) in the opposition’s final third (area of pitch) that is successful (outcome) will have a positive effect on a player's rating.


So there you have it. England look steady across the board, while Italy have notable under- and over-performers.

🧐 What to look out for on Sunday

Possession vs aggression

Italy tired against a ball-hogging Spain side and England will want to adopt a similar game plan here. By the end of that semi-final, Italy were desperate for penalties.

England have had the second-highest number of 10+ pass sequences in the tournament (110), after Spain (176!). England also average 54% possession, the fourth-highest average at Euro 2020. If they can marry Spain's game plan with clinical finishing, they'll be onto a winner.

Italy will press high. They have won 51 high turnovers (4th highest at the tournament) and 12 of these have ended in a shot (highest in tournament), with 3 (also a high) ending in a goal. A high turnover just means pressing and winning the ball in the attacking third.

Jordan Pickford looked shaky in the semi-final with the ball at his feet and you know Italy will look to exploit his kicking. Expect them to block off his short passing routes, force errors through intense pressing, and attack immediately once the ball is won. Denmark squandered opportunities that Italy would take.

Were I Italy (and one day, I might be), I’d try to cut off Pickford’s route to John Stones in particular. Stones has a pass completion rate of 94% - the highest of any player on the pitch for more than 300 minutes at the whole tournament.

Pickford’s not a tall goalkeeper (1.85m) either, so the accurate shooting of Federico Chiesa could reap rewards if they can get it to him quickly. Italy have 3 goals from outside the box so far; England have 0.

On the other hand (and maybe you won’t find this as intriguing as I did), England have taken 11% of their shots from within the opposition’s 6 yard box. In comparison, 2% of Italy’s shots have been taken from this area.

Here’s the kicker: 7% of the shots against Italy so far have been from within the 6 yard box. They sit a little deeper and give the opposition that chance. Bear in mind that these opportunities have come against much weaker teams than England.

And as if this couldn’t get more amazing, England are yet to be on the receiving end of a shot within their own 6 yard box in Euro 2020. Mamma mia!

Both teams will therefore be able to implement the same attacking plans that have served them so well up to this point.

I think Italy will go for a few very intense pressures early on to unsettle Pickford. England will go for a couple of fast, direct attacks to try and score a nerve-settling goal. If both fail, we’ll settle into a more familiar pattern with Italy choosing their moments to press.

🏎 Feed the speed

But, if England can beat that high press (and they will know it’s coming), they have the speed to cut through the Italian defence. If Italy gamble and commit to the aggressive press, they will be light in numbers at the back. Expect “tactical” fouls as England try to break. (2 of Italy’s 7 yellow cards are classified as “Other” and they really could be for anything.)

Bukayo Saka is expected to start and Kane will look to find him with through balls behind Italy’s back line, as he did so successfully against Denmark. Chiellini and Bonucci have been excellent again though, so England will have to be crafty to outwit them.

Bonucci has the third-highest number of interceptions in the tournament (12), and Italy have faced no successful through balls so far. England will want to test that peculiar statistic, but it is surely a result of Italy’s defensive setup.

Fast, short passing might be a more effective route to goal for England than direct passing and Foden/Grealish would offer this option. Expect long diagonals from Phillips out to the wings, then shorter passing to work into the box.

England have 10 fewer offsides than Italy so far (13 to 23), which seems unexceptional but is quite striking. With VAR ready to pounce on the slightest transgression, I wouldn’t be surprised if England have worked on this. Or at least talked about it briefly.

🤕 Heading for glory?

England have 5 headed goals at the tournament - 2 more than anyone else. England will therefore try to win set pieces and look to exploit their height advantage. Actually, they’ll just aim for Harry Maguire. He wins an average of 4.3 aerial duels per game; Chiellini wins 1.5.

This is where Saka, Sterling, Grealish, and/or Foden can be very useful. Their fancy footwork and slick passing will move the Italian defence out of position, and that’s not where they like to be.

England goals at Euro 2020

Gareth Southgate may prefer to use Grealish/Foden as subs again, once the opposition starts to tire both physically and mentally.

Italy will turn to Domenico Berardi off the bench. He has a tournament high of 7 shot-ending carries (he has dribbled and then shot 7 times, basically) and has only played 309 minutes. Admittedly, none of those 7 has been on target so far.

This is a theme with Italy. They have hit twice as many shots off target as England (50 vs 25).


👊 Key duels

Phillips vs Verratti

Verratti is a sublime passer, but he won’t get much time on the ball with Leeds United’s Kalvin Phillips around.

Those are pretty astonishing numbers, right? Kalvin ran over 15km in the semi-final, too.

Note that Harry Kane drops into midfield to press the other team. This is where Italy really build their attacks, so England will prioritise the middle third again. They move the ball much more quickly than past Italy teams, however.

Sterling moves into prime goalscoring positions, waiting for the turnover. He has only applied one pressure in the attacking third at the whole tournament. Phillips literally does the work of multiple players in this regard.

There’s more to Verratti’s game than passing, though. He has made more tackles (21) than anyone else in the tournament - and he missed the first couple of games. He has created 10 chances, second only to Jordi Alba. If England can cut off his service (easier said than done), they will sever the most important link between Italy’s midfield and attack.

Locatelli (14) is second on the list for tackles made at Euro 2020, while Jorginho has the third-highest number of ball recoveries (40) in the tournament. Jorginho also has 21 interceptions, 7 more than any other player. England would be well advised to skip this part of the pitch, using longer passes to the wings.

I’d expect Italy to start with Barella over Locatelli though, especially as the former brings much-needed speed to the Italian midfield. He hasn’t shown his Inter form yet, but Barella is a real threat in the opposition box and I could see him popping up with a goal here. Especially if Phillips is preoccupied with Verratti.


Chiellini vs Sterling

Speaking of speed… Chiellini as not as slow as you think. He has reached a handy 29 km/h in the tournament, but Sterling (33.1km/h) is one of the 10 fastest players at Euro 2020.

Acceleration is the key though and Sterling will want to vary the pace, as he did to great effect against Denmark.

Sterling also has the highest Expected Goals (xG) of any player at Euro 2020 with 3.92. That just means that based on the quality of chances he has had, he’d be expected to score 3.92 goals on average. He has scored 3, so he is underperforming a little based on that metric. England will forgive him, I’m sure.

Kane is third on the list (Morata is second…), but Kane has 4 goals from an xG of 3.57. This total does not include penalties. It includes penalty rebounds, though.

🤯 And get this: Ciro Immobile (2.24 xG) is the only Italian player in the top 30 at Euro 2020, based on expected goals. He has taken 17 shots, the third highest total in the competition. He must just be shooting from the halfway line.

Spinazzola (1.14) is Italy’s next entrant at 34th on the xG list, and he’s injured.

Italy shoot from further out, so statistically they are less likely to score from these positions. Yet they have scored more goals than England. Statistically, that trend can’t go on forever - but maybe just one more match before we regress to the mean?

Chiesa vs Shaw

Italy should put Chiesa up against Luke Shaw, rather than Kyle Walker. Shaw has 2 assists and an Expected Assist (xA) total of 1.64, meaning he has passed to players in strong shooting positions. While he’s doing all that attacking, he is leaving space in defence.

Kyle Walker has been left in defence to cover, due to his exceptional pace. He has an xA of just 0.14, but has 33 ball recoveries - second only to Kalvin Phillips (37) in the England team. Walker has been superb at stifling counter-attacks and he’ll have a job on this Sunday.

So, Italy will get more joy by pitting Chiesa against Shaw rather than Walker. At the other end, Shaw will fancy his chances against Di Lorenzo, who has committed 11 fouls (second-highest at Euro 2020). With Spinazzola replaced by the weaker Emerson on the left of Italy’s defence, England should see Italy’s wings as areas to exploit.

Spinazzola was such an important player for Italy and I fear his absence could make the difference. He averaged 1.8 successful dribbles per game, twice as many as any other Italian player. 28% of Italy’s shots have come from the left and, while Emerson picked up that mantle in the semi-final, he was much less accurate than Spinazzola.

And look, if I can’t engage in tired national stereotypes I just won’t do this newsletter any more.

Again we see that England shoot from the centre of goal, and they do so from close to the goal.

This is very much the Pep Guardiola approach that has worked so well at Man City. Why bother shooting from places where you’re unlikely to score? (That’s rhetorical: The real reason is because it’s fun to shoot from miles out.)

But Italy will crowd that part of the pitch, so England could use Kane to occupy the centre-backs and work Sterling into the gap between Emerson and Chiellini.


What I’d bet on, were I of the gambling community

  • Federico Chiesa to have 1+ shots from outside the box.

  • Marco Verratti to receive a yellow card. And then do this: 🤌

  • Raheem Sterling to have 2+ shots on target.

  • Harry Maguire to have 1+ shots on target.

  • Kalvin Phillips to make 4+ tackles.

  • Harry Kane to score.

  • Under 8.5 corners. (Italy have conceded an average of 3.83 corners per match compared to England’s 3.5.)

  • Both teams to score.

I would NOT put all of these into an accumulator, you lunatic.

🔮 Prediction

(You just skipped to this bit, didn’t you?)

🇮🇹 1-1: Italy to win on penalties.

I have to stick with Italy, don’t I? If the game is tense, it will suit them very well. They will grind out the win however they need to.

Wherever and however you are watching, do enjoy the game. And if England win, please don’t set off fireworks near my house.

Ciao! 👋

Loading more posts…