Your teams should be working closer to the web platform with a lot less complex abstractions. We need to relearn what the web is capable of and go back to that.
Let’s be clear, I’m not suggesting this is strictly better and the answer to all of your problems. I’m suggesting this as an intentional business tradeoff that I think provides more value and is less costly in the long run.
The whole point of the web is that we’re not supposed to be dependent on any one company or person or community to make it all work and the only reason why we trusted Google is because the analytics money flowed in our direction. Now that it doesn’t, the whole internet feels unstable. As if all these websites and publishers had set up shop perilously on the edge of an active volcano.
But that instability was always there.
This is the transcript of a fantastic talk called “The Tools We Still Need to Build with AI.”
Absorb every word!
The rise of dot-com companies was pitched as a no consequences gold rush. We were on the precipice of a fictional future where everyone would be cashing in on the web. The reality was quite a bit more slow, and boring. Business on the web consolidated, as we now know, and left most people holding the bag. There’s no knowing exactly what will happen with AI technologies, but it wouldn’t be unreasonable to expect something far more boring and centralized than what’s being promised.
Primer was a film about a start-up …and time travel. This is a short story about big tech …and time travel.
I find my feelings about AI are actually pretty similar to my feelings about blockchains: they do a poor job of much of what people try to do with them, they can’t do the things their creators claim they one day might, and many of the things they are well suited to do may not be altogether that beneficial. And while I do think that AI tools are more broadly useful than blockchains, they also come with similarly monstrous costs.
A very even-handed take.
I’m glad that I took the time to experiment with AI tools, both because I understand them better and because I have found them to be useful in my day-to-day life. But even as someone who has used them and found them helpful, it’s remarkable to see the gap between what they can do and what their promoters promise they will someday be able to do. The benefits, though extant, seem to pale in comparison to the costs.
Powerful metaphors in this piece by Maria Farrell and Robin Berjon on the Waldsterben of the internet:
Our online spaces are not ecosystems, though tech firms love that word. They’re plantations; highly concentrated and controlled environments, closer kin to the industrial farming of the cattle feedlot or battery chicken farms that madden the creatures trapped within.
We all know this. We see it each time we reach for our phones. But what most people have missed is how this concentration reaches deep into the internet’s infrastructure — the pipes and protocols, cables and networks, search engines and browsers. These structures determine how we build and use the internet, now and in the future.
I don’t believe the greatest societal risk is that a sentient artificial intelligence is going to kill us all. I think our undoing is simpler than that. I think that most of our lives are going to be shorter and more miserable than they could have been, thanks to the unchecked greed that’s fed this rally. (Okay, this and crypto.)
I like this analogy:
AI is like a dancing bear. This was a profitable sideshow dating back to the middle ages: all it takes is a bear, some time, and a complete lack of ethics. Today, our carnival barkers are the AI startups and their CEOs. They’re trying to convince you that if they can show you a bear that can dance, then you’ll believe it can draw, write coherent sentences, and help you with your app’s marketing strategy.
Part of the curiosity of a dancing bear is the implicit risk that it’ll remember at some point that it’s a bear, and maul whoever is nearby. The fear is a selling point. Likewise, some AI vendors have even learned that the product is more compelling if it’s perceived as dangerous. It’s common for AI startup execs to say things like, “of course there’s a real risk that an army of dancing bears will eventually kill us all. Anyway, here’s what we’re working on…” How brave of them.
It’s been dispiriting but unsurprising to see American commentators weigh in on the EU’s Digital Markets Act. I really wish they’d read Baldur’s excellent explainer first.
John has been doing his predictable “leave Britney alone!” schtick with regards to Apple (and in this case, Google and Facebook too). Ian Betteridge does an excellent job of setting him straight:
A lot of commentators seem to have the same issue as John: that it’s weird that a governmental body can or should define how products should be designed.
But governments mandate how products are designed all the time, and not just in the EU. Take another market which is pretty big: cars. All cars have to feature safety equipment, which varies from region to region but will broadly include everything from seatbelts to crumple zones. Cars have rules for emissions, for fuel efficiency, all of which are designing how a car should work.
But there’s one assumption in John’s post that Ian didn’t push back on. John said:
It’s certainly possible that Meta can devise ways to serve non-personalized contextual ads that generate sufficient revenue per user.
That comes with a footnote:
One obvious solution would be to show more ads — a lot more ads — to make up for the difference in revenue. So if contextual ads generate, say, one-tenth the revenue of targeted ads, Meta could show 10 times as many ads to users who opt out of targeting. I don’t think 10× is an outlandish multiplier there — given how remarkably profitable Meta’s advertising business is, it might even need to be higher than that.
It’s almost like an article of faith that behavioural advertising is more effective than contextual advertising. But there’s no data to support this. Quite the opposite. I wrote about this four years ago.
Once again, I urge you to read the excellent analysis by Jesse Frederik and Maurits Martijn.
There’s also Tim Hwang’s book, Subprime Attention Crisis:
From the unreliability of advertising numbers and the unregulated automation of advertising bidding wars, to the simple fact that online ads mostly fail to work, Hwang demonstrates that while consumers’ attention has never been more prized, the true value of that attention itself—much like subprime mortgages—is wildly misrepresented.
More recently Dave Karpf said what we’re all thinking:
The thing I want to stress about microtargeted ads is that the current version is perpetually trash, and we’re always just a few years away from the bugs getting worked out.
The EFF are calling for a ban. Should that happen, the sky would not fall. Contrary to what John thinks, revenue would not plummet. Contextual advertising works just fine …without the need for invasive surveillance and tracking.
Tracker-driven behavioural advertising is bad for users. The advertisements are irrelevant most of the time, and on the few occasions where the advertising hits the mark, it just feels creepy.
Tracker-driven behavioural advertising is bad for advertisers. They spend their hard-earned money on invasive ad tech that results in no more sales or brand recognition than if they had relied on good ol’ contextual advertising.
Tracker-driven behavioural advertising is very bad for the web. Megabytes of third-party JavaScript are injected at exactly the wrong moment to make for the worst possible performance. And if that doesn’t ruin the user experience enough, there are still invasive overlays and consent forms to click through (which, ironically, gets people mad at the legislation—like GDPR—instead of the underlying reason for these annoying overlays: unnecessary surveillance and tracking by the site you’re visiting).
The EU is not the FCC. I wish every American tech pundit would read and digest this explainer before writing their thinkpieces.
It’s very common for US punditry to completely misunderstand the EU and analyse it as if it were a US political entity – imagining that its actions are driven by the same political and social dynamics as a protectionist industry within the US.
It seems to me that there is a fundamental discrepancy between the way readers interact with books and the way the hack-your-brain tech community does. A wide swath of the ruling class sees books as data-intake vehicles for optimizing knowledge rather than, you know, things to intellectually engage with.
In a world where tech billionaires dominate so much of our culture, it’s troubling to see books treated like mere vessels for self-betterment the way that cold-water therapy and Fitbits are. Some of us enjoy fiction.
Ever wondered why you’re always being encouraged to download the app?
But zero percent of app users have installed an ad-blocker, because they don’t exist, because you’d go to prison if you made one. An app is just a web-page wrapped in enough IP to make it a felony to add an ad-blocker to it.
I understand that OpenAI/Microsoft can’t build ChatGPT within our legal framework. Well they could but it would be prohibitively expensive (it already is now without paying the people who did the work). But I missed the part where that is our problem as a society.
This!
I am tired of talking about these things as tech issues. They are not. They are social and political.
I agree with everything that Matt says here. Evangelising accessibility by extolling the business benefits might be a good strategy for dealing with psychopaths, but it’s a lousy way to convince most humans.
The moment you frame the case for any kind of inclusion or equity around the money an organization stands to gain (or save), you have already lost. What you have done is turn a moral case, one where you have the high ground, into an economic one, where, unless you have an MBA in your pocket, you are hopelessly out of your depth.
If you win a business-case argument, the users you wanted to benefit are no longer your north star. It’s money.
There are a lot of astute observations in here.
This rings true to me.
Google has a serious AI problem. That problem isn’t “how to integrate AI into Google products?” That problem is “how to exclude AI-generated nonsense from Google products?”
Web performance is an unalloyed good. No one has ever complained that a website is too fast.
So the benefit is pretty obvious. Users like fast websites. But there are other benefits to web performance. And they don’t all get equal airtime.
A lot of good web performance practices come down to the first half of Postel’s Law: be conservative in what send. Images, fonts, JavaScript …remove what you don’t need and optimise the hell out of what’s left.
That can translate to savings. If you’re paying for the bandwidth every time a hefty file is downloaded, your monthly bill could get pretty big.
So apart from the indirect business benefits of happy users converting to happy customers, there can be a real nuts’n’bolts bottom-line saving to be made by having a snappy website.
This is related to the cost-savings benefit. If you’re shipping less stuff down the wire, and you’re optimising what you do send, then there’s less energy required.
Whether less energy directly translates to a smaller carbon footprint depends on how the energy is being generated. If your servers are running on 100% renewable energy sources, then reducing the output of your responses won’t reduce your carbon footprint.
But there’s an energy cost at the other end too. Think of all the devices making requests to your server. If you’re making those devices work hard—by downloading, parsing, executing lots of JavaScript, for example—then you’re draining battery life. And you can’t guarantee that the battery will be replenished from renewable energy sources.
That’s why sites like the website carbon calculator have so much crossover with web performance:
From data centres to transmission networks to the billions of connected devices that we hold in our hands, it is all consuming electricity, and in turn producing carbon emissions equal to or greater than the global aviation industry. Yikes!
There comes a point when a slow website isn’t just inconvenient, it’s inaccessible.
I’ve always liked the German phrase for accessible: barrierefrei—free of barriers. With every file you add to a website’s dependencies, you’re adding one more barrier. Eventually the barrier is insurmountable for people with older devices or slower internet connections. If they can no longer access your website, your website is quite literally inaccessible.
I’ve noticed that when it comes to making the argument in favour of better web performance, people often default to the business benefits.
I get it. We’re always being told to speak the language of business. The psychology seems pretty straightforward; if you think that the people you’re trying to convince are mostly concerned with the bottom line, use the language of commerce to change their minds.
But that’s always felt reductive to me.
Sure, those people almost certainly do care about the business. Who doesn’t? But they’re also humans. I feel like if really want to convince them, speak to their hearts. Show them the bigger picture.
Eliel Saarinen said:
Always design a thing by considering it in its next larger context; a chair in a room, a room in a house, a house in an environment, an environment in a city plan.
I think the same could apply to making the case for web performance. Don’t stop at the obvious benefits. Go wider. Show the big-picture implications.
Baldur has new book coming out:
The Intelligence Illusion is an exhaustively researched guide to the business risks of Generative AI.
Design Thinking didn’t change business at all, rather it changed Design into business, adopting its language, priorities and techniques. It sold out Design in an attempt to impress those in power, and in so doing lost its heart.