The year that was in tech

Episode Summary

Executive Summary: In its 2023 year-end review, BBC Tech Life examines generative AI’s promise and risks, including regulation, jobs, security and environmental costs. The programme also highlights practical technology in Ukraine and Nigeria, Kenya’s digital economy and electric transport, the human toll of content moderation, and experiments from remote-controlled cars to AI-powered wildlife deterrents.

Main Topics: AI’s rapid growth, risks and regulation: The programme traces generative AI’s mainstream arrival, from creative tools such as Adobe’s to concerns about job displacement, fraud, discrimination and concentration of economic value. Guests argue that governance and human-centred design matter; EU lawmakers agreed on comprehensive AI rules, expected to apply in 2026. Technology’s environmental footprint: AI depends on energy-intensive data centres, cooling water and resource extraction. A proposed Google data centre in drought-stricken Uruguay drew public anger over water use, while author Kate Crawford described the broader costs in terms of energy, data, labour and minerals. Technology and everyday resilience: The programme showcases 3D printing for rebuilding schools in Ukraine and digitising Nigerian newspapers to preserve national history. It also reports how a cyberattack disrupted Kenya’s e-Citizen services and mobile-money payments, illustrating the risks of dependence on digital infrastructure. Kenya’s innovation and electric mobility: Reporting from Nairobi spotlights entrepreneurs using digital services, blockchain and AI to address financial access and agricultural crop loss. Electric-bus maker BasiGo sees an opportunity to decarbonise transport by drawing on Kenya’s renewable electricity supply. Autonomous vehicles and trust: The show reviews a difficult year for autonomous driving, including regulatory interventions and Tesla recalls. Remote-controlled vehicles offer a possible intermediate step: people may trust driverless-looking cars more when they know a human is controlling them. Human consequences and unexpected uses of tech: A former Kenyan content moderator describes lasting psychological harm from viewing violent material and the distress of losing his job. Other stories include a wealthy technologist pursuing life extension and a homeowner using machine learning to deter badgers and foxes.

Key Arguments: Generative AI is already accessible to the public, so the practical challenge is not whether it will arrive but how its uses and harms will be managed. AI risks include malicious deception, discrimination and unequal distribution of economic gains, alongside longer-term questions about whether models consistently serve human interests. AI’s environmental costs are broader than electricity consumption: data collection, hidden human labour, cooling water, and mineral extraction all contribute to its footprint. Digital services can expand access and improve livelihoods, but cyberattacks can leave people unable to use essential public services or make payments when societies rely heavily on them. Technology can help address local problems, including crop losses and emissions, but its benefits depend on appropriate infrastructure, reliable implementation and sustainable resource use. Human oversight may help build public trust in automated transport, while the experience of content moderators demonstrates that technology’s harms can also fall heavily on workers behind the scenes.

Data Points: Expected application of EU AI rules: 2026 - The programme says the first comprehensive EU AI legislation was expected to enter into force the following year, with its rules applying in 2026. Possible NVIDIA AI-server electricity use: Equivalent to half of current global data-centre electricity consumption - Academic Alex de Vries describes this as a possible level by 2027 if chip supply constraints ease and demand remains high. Possible AI-server share of global electricity consumption: About 0.5% - The programme gives this as the estimated global electricity share associated with the AI servers discussed. Proposed Google data-centre water use: More than 7 million litres per day - A study obtained by an academic in Uruguay reportedly estimated this amount for cooling, with water sourced from public drinking-water supplies; Google said its plans were exploratory and being revised. Kenyan use of mobile money: Over 57% - Reporter Anita Nkonji cites this share when explaining how disruption to mobile-money payments left people unable to transact. Services on Kenya’s e-Citizen portal: Over 5,000 - The portal provides government services including passports, e-visas and driving licences; it was reportedly inaccessible during a cyberattack. Avocado production and loss in Kenya: About 30% of the population grows avocados; estimated losses of 40–45%, or 30,000–40,000 metric tons per year - An entrepreneur describes the scale of post-harvest losses and a blockchain-and-AI system intended to improve coordination and farm advice. Kenya’s electricity from renewable sources: Nearly 90% - BasiGo’s Moses Dorito cites the renewable share as an opportunity for electric buses; he also describes full reservations for the company’s possible 2024 deliveries and taking reservations for 2025. African buses on the roads: Over 1 million - BasiGo’s CEO cites this as the potential scale of the continent-wide opportunity to decarbonise bus transport. Time to 3D-print a school building: Three weeks - A project representative in Ukraine says a classroom structure was printed in three weeks, with faster construction anticipated as the process scales.

Pivotal Quotes: "There's no point all of us sitting here going, oh, AI is going to destroy the world. Well, it's happening." — Martha Lane Fox: She argues that people should engage seriously with AI’s effects rather than assume the technology can simply be stopped. "So, we are really looking at an enormous extractive industry for the 21st century." — Kate Crawford: She summarises AI’s dependence on data, hidden labour, electricity, water and mined resources. "The sacrifices I made, I sacrificed my human side for this job. I don't think you can give any more than your soul." — Trevin Brownie: The former content moderator describes the personal cost of viewing disturbing online content and the impact of being fired.

Implications: AI’s benefits will depend on effective oversight and attention to workers, security and unequal outcomes, as well as energy, water and mineral costs. Meanwhile, digital tools can solve local problems, but critical services need resilience and technology workers need stronger protections.