How Did We Get Here?
These are my notes for my talk at Wuthering Bytes 2023 in Hebden Bridge, on August 25. It builds on a lot of earlier stuff, most significantly a talk I gave at OpenTech in.. checks notes.. 2013.
You can read the OpenTech version here:
They were, in hindsight, days of innocence and optimism https://www.opentech.org.uk/2013/
Bill Thompson
Song
And you may find yourself living on a protocol stack
And you may find yourself on neglected subnet
And you may find yourself designing the screen of a new social tool
And you may find yourself on a functional web, with no context collapse
And you may ask yourself, “Well, how did I get here?”
Watching the packets go by, let the voltage hold me down
Watching the packets flow by, data routed back to source
Thanks to the BGP, after the TTL’s gone
Once in a lifetime, packets flowing round and round
And you may ask yourself, “How do I code this?”
And you may ask yourself, “Where is that new style sheet?”
And you may tell yourself, “This is not my beautiful net
And you may tell yourself, “This is not my beautiful web”
Acknowledgement
Before we begin, I want to acknowledge that this debate happens in the shadow of a climate crisis that has burned Rhodes and Hawaii and Canada, flooded Pakistan, killed infant penguins in Antarctica, destroyed towns and cities and lives, and threatens to overwhelm us.
I’m not going to claim that a healthy digital public sphere can offer us answer, particularly since these technologies are themselves currently contributors to global heating,but I do think both that the toxic network doesn’t help and that a better net might help us find a way forward, if only by amplifying human connections and cognitve capacity. It seems worth trying.
What I plan to talk about
From TCP/IP to Instagram Threads, a journey into the betrayal of our dreams and a glimpse of redemption.
I’ve been on the internet since 1985 and organised the 1986 Community Computing Network conference via X.25 dialup using Geonet and the Manchester Host.
I was at the first World Wide Web conference in 1994, built websites for Comic Relief, the Edinburgh Fringe and my local MP. Now I work for BBC Research & Development where my job is to ‘fix the internet’.
I’ve watched the promise sour, seen the tools I championed and helped develop used to control, damage, and oppress. And I’ve seen core principles of open source, open data and an open web used by those I’d consider enemies to achieve goals I abhor, along with giving some people and companies incredible wealth and political power.
From IRC to ICQ, MySpace and Friendster, Friends United, Amazon, Facebook,Twitter, Signal, Threads and X, and from algorithms to generative AI, Palantir and OpenAI, technological advances over the past decade have profoundly changed society.
Many of these changes have been positive, connecting people in new ways, expanding new forms of entrepreneurship, helping government and cities run more efficiently, and improving health and medical care.
The network and its many applications have had a massive impact on our lives, but the tools they gave us were unable to resist the incursions of predatory capitalism or hold back the worst excesses of human toxicity and hatred.
People are suffering. Our institutions are suffering. And the processes of democracy are suffering, which should worry us as next year we will see elections in South Africa, Ghana, Tunisia, Mexico, India, Austria, Belgium, Lithuania, Moldova and Slovakia, as well as the UK.
Oh, and there’s the US.
X and Facebook and TikTok won’t decide the results, but they have changed the way things work, given advantage to new operators and forms of work, and – with the boost offered by generative AI tools – it doesn’t look good for ‘free’ or ‘fair’.
We are now faced with an almost unimagibly difficult task, but we can’t just give up, because we deserve better – everyone whose data is appropriated for targetting advertising or training a model to be used to oppress them, everyone who is attacked, every trans person facing the weight of organised hatred, every single one of us.
Where did it go wrong?
If I was asked to point to one cause, I’d say we are where we are because allowed ourselves to develop a whole load of technologies based on the blissfully optimistic assumptions of a group of white men like me who either believed that the world thought like they did, or didn’t bother to consider the question.
Maurice Wilkes
David Whitehead
Developed the stored programme digital computer
Vint Cerf
Bob Kahn
Gave us TCP/IP
Jon Postel
Created the Internet Assigned Numbers Authority which became ICANN
Tim Berners Lee
Inveted HTML and HTTP
Mark Zuckerberg
Jack Dorsey
Took nascent social networks and turned them into platforms
Karen Spärck Jones
Geoff Hinton
Develope NLP and AI
And now of course Elon Musk
And we have not listened to other voices
Like Timnit Gebru, Margaret Mitchell, Emily Bender
Like Brenda Laurel, many years ago
But we are where we are.
We’re not going to turn the network off and decide to live offline. It makes no sense to ask whether we need the Internet because in the world as it is the internet is part of the lived reality, not just in the west but everywhere with a mobile signal.
We Have Choices
The core technology behind the network is two-fold: a way of moving data between computers, and a language for those computers to speak.
The language spoken evolved from the network control protocol to the transmission control protocol and the internet protocol, a digital esperanto devised in 1974 by Vint Cerf and Bob Kahn and still the way that my laptop and even your home television manage the flow of bits that offer emails and messages and sound and vision.
We built it all. Engineers, mostly, working initially to provide a robust, fast, efficient data network that would support scientific research, government activity and military command and control but now focused on entertainment and the effective delivery of advertising.
Businesses came late to the party. Ordinary users later still, and we found that most of the big decisions had already been made. The network we were allowed to join was based around open standards, so anyone could write programs that used it. It did not require any authentication of identity before you could join, just an allocated identification number or IP address that belonged to an organisation rather than an individual. Information was not logged or tracked and could not easily be monitored in transit.
It was, in short, a digital expression of the philosophy of the West Coast liberal intelligensia who built it. It served the needs of a community where most people knew each other, where simplicity was more important than control, and where openness was not just a virtue but an overriding ideology.
But what can be built can be rebuilt, and assumptions can be overturned. Anyone who, like me, lived through Thatcherism knows just what can be achieved by someone who is willing to destroy all established institutions in pursuit of their goal.
So it is with the net today. The underlying technology is more and more based around architectures of control.
If there is freedom, it is the freedom to engage on the platform’s terms, to be surveilled in pursuit of commerce and social coercion, and to have no agency other than that which serves the platform-governmental complex.
It’s not a perfect alignment – the fights over news in Canada and Australia, the UK government’s attempt to legislate for Pi to be an rational number – or its equivalent, to put backdoors in end to end encryption and have it remain secure – or the Chinese government’s work to keep Baidu and Tencent in check are some examples.
But overall, the interests of governments and companies are far more aligned with each other than those of the people using the network and the democratic deficit in network governance is real and not being seriously addressed. If you think of AI systems as being the latest layer of the network, additional elements in the computational infrastructure that defines our world, then its notable that all the summits involve governments, businesses, a few academics – and nobody from civil society.
This is a problem, because we need to be in the room. The core principles many of us espouse – free software, open standards, open data, an open internet capable of supporting an open society – leave us weak and exposed when we come up against those who would expropriate our work, use the force of copyright and patent law against us, and turn our co-creators into sharecroppers on their social media platforms.
Thirty years after CERN agreed with TimBL’s request to release the codebase and standards that define the Web, the start of the modern online era, we have Elon Musk turning Twitter into a safe space for extreme views, pushing the technological determinist agenda that would see him and his friends left to run the world as they wish.
And we have no protection, like Earth facing a Kardashev Type III civilisation. He owns the company, he can do these things, and the underlying technologies mean that we can’t implement our own user blocks or other features.
The Open Internet and Its Enemies
We made the network, we can change it. Its future is not decided, it not innate.
For Karl Popper the real danger to open societies lay in the belief that some characteristics of human society are innate, essential and unchangeable, what he terms historicism. He saw the danger of assuming that we could predict the future of society, not least that when things begin to deviate from the path more and more extreme measures are taken.
The techno-determinists may argue that the future development of the network towards the support structure for capital is self-evident. We know otherwise.
But that means that we have a responsibility for what the internet is, has been and may become that we cannot escape. We built this thing, our actions shaped it, and we are accountable for what it does.
They stole our revolution – now we want it back was NTK’s slogan twenty-five years ago
Google took HREF
Facebook took HTML
OpenAI took ALL THE CONTENT
What’s going to be lost next? Is the liminal space doomed to be another zone of commercial exploitation, as the web and mobile ecosystems have proven to be? Are those of us toiling in the public service, or trying to build open and adaptable systems, merely clearing the forest and tilling the land so that GM crops can be planted?
We can put social justice at the heart of everything we build
We can consider the social impact of what we do
But it will not, in itself, stop our ideas and works being abused, stolen and apprpriated.
We’re like The Clash in 1977, watching our fine friends in new boots[1].
So how can we ensure that the network is defended? Or even defendable.
Network Evolution
We can start by creating alternatives. After some years of stasis we are living through a Cambrian explosion in new media forms and voices, the online equivalent of that amazing period 540 million years ago when the fossil record shows the sudden – geologically speaking – appearance of many of the ancestors of modern species.
The Cambrian was a period of great experimentation in structure and function, but many promising models simply died out, leaving the world to our ancestors – four limbs, bilateral symmetry. There is no good reason to suppose that other forms of life were, in themselves, unable to be sustained if conditions – chemical, physical – had been different or if they had simply been luckier.
So it is today.
Experiments are being tried all over the net. Some will succeed and shape the future of media. Some will fail. Many approaches – and the communities built around them will crash and burn, and most of the experiments will fail, but there will be immense evolutionary pressure and we will quickly see better ways emerge
Similarly, governments and societies may decide to embrace the new models, but some will resist and some will make the wrong choices. Those that succeed may shape the environment and make it more likely that others will fail, just as YouTube’s success has polluted the ecosystem and shrunk the niche for other video-sharing sites.
Towards a New Luddism
Throughout history new technologies have been pressed into the service of the dominant classes throughout history and consider how the systems we are building, the datasets we’re liberating, the Fediverse we’re enabling, and the principles we espouse can be used to increase social justice, reduce inequality and oppression, and serve the mass of humanity not just the wealthy superclass.
It’s worth reflecting on the Luddites.
The Luddites were not opposed to technology, and used some of the most advanced cloth-making technologies of their day. They drew support from skilled workers, with some degree of autonomy.
They revolted because the new forms of automation introduced in the early 19th century were amenable to industrial organisation and required such an investment of capital to buy and install that factory working was the only way to make them profitable.
They fought because the capitalist mill owners used their political powers to change the rules to their advantage, reducing the power of the guilds over the ways cloth and wool were made and manufactured, just as rules on personal privacy are in danger of being diluted in favour of social network service providers.
They lost. How can we do better?
Perhaps we can be inspired by a current example of a group of skilled workers refusing simply to acquiesce to the demands of capital. Not spinning up ActivityPub servers, though that does matter, but the writers and actors strikes in the US at the moment, prompted in part by resistance to having the capabilities of AI systems used to devalue their work, on script and on screen.
On July 14, 2023, the American actors’ union SAG-AFTRA went on strike over an ongoing labor dispute with the Alliance of Motion Picture and Television Producers. It has coincided with a Writers Guild of America strike in a series of broader Hollywood labor disputes.
Refusing to accept that the owners of the technology get to dictate the terms to the workforce
Like the Luddites, they can see that the result will be a greater concentration of power, a loss of control for them, reduced rewards and a poorer product for consumers.
As I said at the start, I don’t think we can, or should, try to live without the Internet. But I do think that we need to seize control of the means of online production and use them for the benefit of the many, not the few, and in ways that we understand and endorse.
In the end the issues facing the world are not directly concerned with technology or the internet but about living together and finding ways to allow each one of us to achieve all that he or she is capable of.
If computers and the network can help to reduce poverty and want, and offer ways to liberate human potential, then we should embrace it –not for its own sake but because it helps us achieve our potential as human beings.
But we need to shape it to these ends, against enormous forces and eighty years of development. The important thing is to start, and keep going
As Tony Benn liked to say
There is no final victory, as there is no final defeat. There is just the same battle. To be fought, over and over again. So toughen up, bloody toughen up.
Same as it never was
And you may find yourself living on a protocol stack
And you may find yourself on neglected subnet
And you may find yourself designing the screen of a new social tool
And you may find yourself on a functional web, with no context collapse
And you may ask yourself, “Well, how did I get here?”
Watching the packets go by, let the voltage hold me down
Watching the packets flow by, data routed back to source
Thanks to the BGP, after the TTL’s gone
Once in a lifetime, packets flowing round and round
And you may ask yourself, “How do I code this?”
And you may ask yourself, “Where is that new style sheet?”
And you may tell yourself, “This is not my beautiful net
And you may tell yourself, “This is not my beautiful web”
Watching the packets go by, let the voltage hold me down
Watching the packets flow by, data routed back to source
Thanks to the BGP, after the TTL’s gone
Once in a lifetime, packets flowing round and round
Same as it ever was, same as it ever was
Same as it ever was, same as it ever was
Same as it ever was, same as it ever was
Same as it ever was, same as it ever was
Networks dissolving and communities collapsing
There’s no safety in the middle of the network
Losing connection, making connection
Remove the danger from your online engagement
Networks dissolving and communities collapsing
Watching the packets go by, let the voltage hold me down
Watching the packets flow by, data routed back to source
Thanks to the BGP, after the TTL’s gone
Once in a lifetime, packets flowing round and round
Watching the packets go by, let the voltage hold me down
Watching the packets flow by, data routed back to source
Thanks to the BGP, after the TTL’s gone
Once in a lifetime, packets flowing round and round
You may ask yourself, “What was that functional app?”
You may ask yourself, “Where does that route table point to?”
And you may ask yourself, “Am I v4, am I v6?”
And you may say to yourself, “My God, what have we done?”
Watching the packets go by, let the voltage hold me down
Watching the packets flow by, data routed back to source
Thanks to the BGP, after the TTL’s gone
Once in a lifetime, packets flowing round and round
Watching the packets go by, let the voltage hold me down
Watching the packets flow by, data routed back to source
Under the DNS, there is UDP under all
Once in a lifetime, packets flowing round and round
Same as it ever was, same as it ever was
Same as it ever was, look where my hand was
IP isn’t holding up, IP isn’t after us
Same as it ever was, same as it ever was
Same as it ever was, same as it ever was
Same as it ever was, same as it ever was (I couldn’t get no rest)
Same as it ever was, hey let’s all count our packets
Here comes the route table
Letting the packets go by (same as it ever was, same as it ever was)
Letting the packets go by (same as it ever was, same as it ever was)
Once in a lifetime, letting the protocol down
Letting the packets go by, networks flowing underground
END
.
[1] I’m sure it’s no coincidence that Apple called it Garageband