SWITCHED ON
The daily technology series nobody asked for but everyone needed
Still Switched On: What Twenty-Five Episodes Taught Us About the Future
We covered artificial intelligence, genetic engineering, brain implants, surveillance capitalism, ransomware, and the end of the world as we know it. Here is what all of it actually means.
Technology does not have a direction. It has momentum. The direction is chosen by people — which means it can be chosen differently. That is the most important thing twenty-five episodes of this series can leave you with.
Yesterday we ended up at the largest possible frame — artificial general intelligence, the technological singularity, the alignment problem, and the question of whether the most consequential invention in human history is being built right now by people who are simultaneously excited and terrified, in a competitive environment that creates strong incentives to proceed faster than wisdom would recommend. Today there is no new technology to introduce. No fresh crisis to document. No emerging framework to scrutinise. Today is the day we step back from the torrent, look at what twenty-five episodes of this series actually adds up to, and try to say something honest about what it means to be a person living through this particular moment in technological history. Bear with me. This one is personal.
01 — The Pattern That Ran Through Everything
If you read every episode of this series — and if you did, first of all, thank you, and second, I am sincerely sorry about what I did to your sleep schedule — a pattern emerged that was never the explicit subject of any single episode but ran underneath all of them. It went like this: a technology arrives with extraordinary promise, is deployed at scale before the consequences are understood, generates harms that were foreseeable in retrospect and largely unforeseen in practice, and is then partially addressed by governance frameworks that arrive years or decades after the deployment, imperfectly fit to the problem they are trying to solve, and enforced with tools designed for a previous era.
This pattern appeared in social media and children's mental health. In facial recognition and consent. In algorithmic hiring and credit scoring. In predictive policing and its feedback loops. In autonomous weapons and the accountability vacuum. In quantum computing and the harvest-now-decrypt-later threat that is already operational. In synthetic media and the collapse of evidentiary trust. In CRISPR and germline editing that produced two gene-edited children before the ethics framework existed. In brain-computer interfaces and neural data rights that have barely been defined. In the orbital debris crisis being created in real time by satellite constellations deployed faster than the regulatory environment can track.
This is not a coincidence. It is a structural feature of technological development in a market economy, where the incentive to deploy is immediate and financial, and the incentive to govern is delayed and political. Understanding this pattern is the most practically useful thing Switched On can give you — because it tells you where to look when the next technology arrives promising to change everything. Ask not just what it can do. Ask what it will do when it scales. Ask who is harmed. Ask who governs it. Ask whether the governance exists yet. The answer to that last question, for most technologies in this series, was no.
02 — What to Actually Worry About
Twenty-five episodes of technology coverage produces, in the attentive reader, a certain ambient anxiety. There is a lot to be concerned about. Distinguishing productive concern from paralysing dread requires some calibration, and I want to try to provide it.
The things most worth worrying about are not the dramatic science fiction scenarios — rogue superintelligence, robot uprisings, gene-edited superhumans. These are possible futures that serious people take seriously, but they are not the present. The things most worth worrying about are the quieter, more immediately operational harms: the algorithmic systems making biased decisions about millions of people's lives right now. The children growing up inside social media ecosystems optimised for engagement rather than wellbeing. The encryption infrastructure that is already being harvested for future decryption. The political systems being destabilised by synthetic media at scale. The orbital environment being degraded by constellation deployment that governance cannot keep pace with. The climate window that is not waiting for the governance of green technology to catch up.
These are not hypothetical risks. They are current conditions. The appropriate response to them is not panic — panic is not a strategy — but it is also not the comfortable dismissal that technology will sort it out, or that the market will find the right equilibrium, or that regulators are on it. The evidence of twenty-five episodes suggests that none of those reassurances are reliable. The appropriate response is informed, persistent, democratically expressed concern. Which requires, as a prerequisite, being informed. Which is why this series existed.
The most dangerous thing about technology's pace is not what it does to institutions. It is what it does to attention. When change is continuous and rapid, the capacity to focus on any single problem long enough to demand accountability for it is the first casualty.
03 — What to Stay Curious About
This series has been deliberately honest about problems, because problems are what the breathless promotional coverage leaves out. But honesty also requires acknowledging what is genuinely exciting, because the dismissive counter-reaction to technological hype — the reflexive scepticism that treats all claims of progress as marketing — is its own form of inaccuracy.
The approved CRISPR therapy for sickle cell disease is real and it is changing lives. The cost reduction in solar power over fifteen years is one of the most significant economic and environmental achievements of the twenty-first century. Reusable rockets have genuinely and permanently altered the economics of access to space, with downstream consequences we cannot fully see yet. mRNA vaccine technology, developed over decades and proved in the COVID-19 pandemic, represents a platform with implications for infectious disease that extend far beyond a single virus. Satellite internet is providing connectivity to communities that terrestrial infrastructure would never have reached economically. Post-quantum cryptography standards have been developed and are being deployed, which means the encryption crisis has a response even if the transition is slow.
The people working on these problems — in laboratories, in regulatory bodies, in civil society organisations, in the unglamorous offices of standards bodies and ethics committees — are doing real work that matters. They are not the story that gets told, because the story that gets told is the billionaire with the announcement and the crisis with the body count. The patient, methodical work of making technology safer, more equitable, and more governed is the story that is happening in parallel, and it deserves at least as much attention as the disruption it is trying to keep up with.
04 — The Question of Agency
The frame that technology happens to people — that it arrives from somewhere, deployed by forces beyond individual influence, and that the appropriate response is either enthusiasm or helplessness — is the frame that benefits the people deploying it most. It is also wrong.
Technology is built by people, funded by people, regulated by people, purchased by people, and used by people. At every one of those points, choices are made. Most of those choices are made by a relatively small number of people with concentrated resources and significant insulation from the consequences of the decisions they make. The concentration of those choices — in a handful of companies, in a handful of jurisdictions, in a handful of research labs — is itself a governance problem of the first order, and one that this series has returned to repeatedly because it underlies everything else.
But the choices are not only made at that level. Regulatory pressure comes from organised public opinion. Legal frameworks come from litigation brought by individuals and organisations who refused to accept that the status quo was inevitable. Product decisions have been changed by user behaviour at scale. The algorithmic amplification of harmful content on social media platforms has been partially — imperfectly, incompletely, but partially — addressed because enough people demanded it loudly enough for long enough that the political cost of inaction exceeded the economic cost of adjustment. This is slow and messy and profoundly unsatisfying compared to the pace at which the problems compound. It is also how change actually happens in democratic societies, and it requires participants.
05 — Why This Still Matters
Switched On started with a robot that did not steal your job and ended at the edge of artificial general intelligence. In between we covered children and social media, synthetic reality and the collapse of evidentiary trust, the right to be forgotten, facial recognition without consent, the metaverse's expensive failure, the algorithm that does not care about your wellbeing, autonomous weapons and the accountability vacuum, quantum computing and the encryption crisis, climate technology and Silicon Valley's limits, CRISPR and the governance gap, brain-computer interfaces and neural privacy, anti-aging and its civilisational implications, lab-grown organs and the transplant waiting list, the new space race and its uncomfortable priority questions, satellite internet and the digital divide, electric vehicles and the infrastructure gap, AI regulation and its three incompatible frameworks, AI art and the copyright crisis, algorithmic bias in hiring and criminal justice, tech monopolies and antitrust, online surveillance and the nothing-to-hide myth, cybersecurity and the ransomware industry, and the singularity question itself.
That is not an exhaustive account of the technological landscape. There are topics this series did not reach — nuclear energy's complicated renaissance, the future of work in an AI economy, spatial computing and what it means to blend digital and physical environments, the ethics of human germline genetic engineering at population scale, the geopolitics of semiconductor supply chains, and dozens more. The list of what matters is longer than any daily series can cover. That is the point. The appropriate response to a world in which technology is transforming everything, faster than any single person can track, is not to follow everything. It is to understand the patterns, develop the questions, and apply them wherever the next headline lands.
The daily technology series nobody asked for but everyone needed is ending its first run. The technology it covered is not ending anything. It is accelerating. The questions it raised have not been answered. They are getting louder.
Stay curious. Stay sceptical. Stay informed. And for the love of everything, turn on two-factor authentication. After twenty-five episodes, that is the most actionable thing I can leave you with.
This has been Switched On. It has been, genuinely, a pleasure.
Switched On was a 25-episode daily technology series covering the ideas, systems, and arguments shaping the digital world. Opinionated. Witty. Occasionally wrong. Always worth the argument.
