The \u2018AI\u2019 doctor will see you now Illustration: Jay Cover\/The GuardianAGI Comics\u2014HF-edits Dr-2 Illustration: Jay Cover\/The Guardian<\/p>\n
\u201cDoes it hurt when I do this?\u201d<\/p>\n
AGI Comics\u2014HF-edits Dr-3 Illustration: Jay Cover\/The Guardian<\/p>\n
\u201cYou seem to have dislocat\u2026\u201d<\/p>\n
A Eye: \u201cNOOOO! The problem is a sprain in the brachial plexus due to you lifting that 10kg carton on Wednesday at 2.58pm and not eating enough blah blah\u201d<\/p>\n
AGI Comics\u2014HF-edits Dr-4 Illustration: Jay Cover\/The Guardian<\/p>\n
\u201cWow, err, thanks\u201d<\/p>\n
In 2035, AIs are more than co-pilots in medicine, they have become the frontline for much primary care. Gone is the early morning scramble to get through to a harassed GP receptionist for help. Patients now contact their doctor\u2019s AI<\/a> to explain their ailments. It quickly cross-checks the information against the patient\u2019s medical history and provides a pre-diagnosis, putting the human GP in a position to decide what to do next.<\/p>\n In face-to-face consultations an AI may listen in the background, weigh the facts of the patient\u2019s case against thousands of medical studies, propose a course of drugs based on the full depth of the latest medical research, which the human doctor could never fully digest. It provides a second opinion for the doctor, who can assess its proposals before deciding how to act.<\/p>\n The public may not be wholly averse to this change. Thirty-eight per cent of people favour using AI to speed up triage in the NHS although 52% prefer humans, citing trust and wanting personal interaction, polling by Ipsos in October 2025 found.<\/p>\n Medical screening becomes highly sophisticated and perhaps a little invasive. Doctors could be sent details of your diet and vital statistics tracked by wearable devices. Smart toilets may even analyse your bowel movements. Medicines could be produced exactly tailored to your body and its needs.<\/p>\n If it all works, diseases are spotted quicker and medicines are more accurately prescribed, but the combination of one AI and one human medic creates a new tension. The best doctors become those most adept at interpreting the outputs of the AIs. Medical schools change their teachings to focus more on managing AI medics, and politicians wrestle with how to overhaul medical regulation and ethics.<\/p>\n The future of AI <\/p>\n The rivals racing to create super-intelligence. This was put together in Design and Development <\/p>\n Harry Fischer and Pip Lev<\/p>\n AI v AI: how lawyers could become a thing of the past Illustration: Jay Cover\/The GuardianAGI Comics\u2014HF-edits Law-2 Illustration: Jay Cover\/The GuardianAGI Comics\u2014HF-edits Law-3 Illustration: Jay Cover\/The GuardianAGI Comics\u2014HF-edits Law-4 Illustration: Jay Cover\/The GuardianAGI Comics\u2014HF-edits Law-5 Illustration: Jay Cover\/The Guardian<\/p>\n Justice is increasingly enabled by AI, although some fear it is taking over. Solicitors preparing for trial have learned to delegate the legwork of unearthing case law and planning arguments to an AI, which proposes the best approach for a barrister to take in court. The current problems of AIs making up case law, as happened at least 95 times in July and August, according<\/a> to trackers, have been ironed out. The new, more robust artificial general intelligence (AGI) systems compress days of work into hours, leaving the human solicitor only to check the AI\u2019s brief to the human barrister.<\/p>\n Next, amid a courts backlog, pressure grows to largely replace barristers too. An experiment is launched allowing adversarial AIs to argue cases in front of a human judge and jury. The results prove compelling. Cases are concluded at a fraction of the cost to the taxpayer and much faster. But numerous miscarriages of justice soon emerge. Campaigners for people wrongly imprisoned begin to demand greater transparency about the inner workings and biases of the AI lawyers.<\/p>\n Having unleashed AGI, governments and companies have to constantly surveil the autonomous systems, employing people to sit at banks of screens, shutting down dangerous behaviour, dispatching good AIs to hunt down bad AIs.<\/p>\n The morning routine Illustration: Jay Cover\/The GuardianAGI Comics\u2014HF-edits Morning-2 Illustration: Jay Cover\/The Guardian<\/p>\n Glasses, wake me up at 7am<\/p>\n AGI Comics\u2014HF-edits Morning-3 Illustration: Jay Cover\/The GuardianAGI Comics\u2014HF-edits Morning-4 Illustration: Jay Cover\/The Guardian<\/p>\n WAIT! \u2026 Were there any eggs in the fridge?!<\/p>\n AGI Comics\u2014HF-edits Morning-5 Illustration: Jay Cover\/The Guardian<\/p>\n He\u2019s asleep now, I can\u2019t ask him.<\/p>\n Illustration: Jay Cover\/The Guardian<\/p>\n He\u2019s clumsy when he hasn\u2019t eaten his breakfast eggs.<\/p>\n Illustration: Jay Cover\/The Guardian<\/p>\n What if he drops me? Replaces me?! There\u2019s a new set of glasses out tomorrow!!!<\/p>\n Wearable AI devices, such as glasses, watches and rings, have become ubiquitous. They function as extra senses, spotting things in our environment that we miss, like a lack of eggs in the fridge, or recording our interactions to remind us later of things we forget. But then they also start to do things for us \u2013 and maybe even worry about not doing a good enough job.<\/p>\n \u201cFirst principle is, you have a bunch of these AI agents that do things for you,\u201d says David Shrier, professor of practice, AI and innovation at Imperial Business School. \u201cThese will be specially programmed AIs that have different areas of expertise and are customised \u2013 adapted to you and your specific needs. So they learn \u2026 what you want and they go and do it for you.\u201d<\/p>\n You might start your day with the help of an information agent. \u201cIt goes out, it curates the articles and when you wake up in the morning and as you are brushing your teeth, it\u2019s reading out to you summaries and interpretation of the news. It can provide some deeper understanding for you.\u201d<\/p>\n This second-guessing based on deep algorithmic familiarity extends to your breakfast. If you have been wearing augmented reality glasses, the AI will have noticed when you opened the fridge that you\u2019ve run out of eggs. When you go downstairs at 7am your phone lights up to say an Amazon drone has delivered you some eggs.<\/p>\n Shrier adds that all of this requires consent, meaning \u201cyou know what you are allowing the AI to understand and know about you, and you get to easily opt out if you don\u2019t want your data shared.\u201d<\/p>\n AI on the farm Illustration: Jay Cover\/The Guardian Illustration: Jay Cover\/The Guardian<\/p>\n Old Macdonald<\/p>\n Illustration: Jay Cover\/The Guardian<\/p>\n had a farm<\/p>\n Illustration: Jay Cover\/The Guardian<\/p>\n AI-AI-O<\/p>\n A farmer\u2019s rounds, checking on the health of livestock, crops, feed supplies and machinery \u2013 which can take from dawn to dusk \u2013 become far less onerous. With cameras and sensors rigged to trees, barns, fence posts and itinerant robots, every farm delivers a torrent of data to help increase productivity and animal welfare. Already, in 2025, an AI model is under development to detect early infections in cows by tracking subtle shifts in their social behaviour. A herd in Somerset is being filmed around the clock to train a model to predict if an animal is in the early stages of mastitis, which affects milk production and is an animal welfare problem. A decade later and with data harnessed from millions of farms around the world, AGI advises not only what to plant and when, but how to build stronger ecosystems, and improve soil health. AGI-powered robots could stalk the fields rooting out weeds and reducing the need for herbicides.<\/p>\n Less work, more play Illustration: Jay Cover\/The Guardian Illustration: Jay Cover\/The Guardian<\/p>\n \u2026 and this was called an office \u2026<\/p>\n Illustration: Jay Cover\/The Guardian<\/p>\n Filled with people and chairs and phones and desks and carpet and water coolers and paper<\/p>\n Illustration: Jay Cover\/The Guardian<\/p>\n This was a board meeting.<\/p>\n Yeah, they look bored!<\/p>\n Sports clubs, live entertainment venues and travel agents boom as AGI transforms work, helping millions of white-collar workers whisk through their tasks leaving time for a new life of leisure \u2013 at least that is one theory.<\/p>\n In the early years of AGI only a few people are made fully redundant by full automation, and most stay in work. For a while people retain the same 40-hour working week and simply get more work done.<\/p>\n At work, AI could act like a \u201cbuddy\u201d or a coach in meetings, says Shrier. It might advise you to go easy on another attender because they look a little stressed.<\/p>\n \u201cDuring your meeting you\u2019re getting this sort of coaching and augmentation that makes you better at interacting.\u201d As the chat ends, the AI sends you a list of next steps to take following this meeting \u2013 and slots it into your project plan. As you get to your desk, your professional AI assistant is bringing you the documents and spreadsheets you need to deliver.<\/p>\n AI-augmented economies could grow sharply. But soon people realise they can afford to work less. The 15-hour week<\/a>, which the economist John Maynard Keynes in 1930 predicted would happen by 2030, becomes a reality. Leisure becomes less about rest and orientates towards creative activities, human-to-human socialising and time caring for children or elderly family members, academics have suggested<\/a>.<\/p>\n
\n\t\tcollaboration with the Editorial Design<\/a> team. Read more from the series.<\/a><\/p>\n