AI cancer cures slowed by chip shortage, says UK's biggest tech boss
Tuesday, 8 September 2026
Warm-up
- How do you think AI can help doctors and patients today?
- What new technology would you like to see in the next five years?
- Should countries try to build more chip factories? Why or why not?
Vocabulary
- modelling
- Building a detailed computer version of something to study how it works.
- marker
- A sign in the body, such as a DNA signal, that shows a condition.
- humanoid
- Shaped or acting like a human being.
- shortage
- A situation where there is not enough of something needed.
- data centre
- A building with many computers that store and process data.
- manufacture
- To make products in large numbers, usually in factories.
- sceptical
- Not fully convinced that something will happen or is true.
- demand
- How much people or companies want to buy a product or service.
Reading
Rene Haas, chief executive of Cambridge-based Arm Holdings, says artificial intelligence will help cure cancer within our lifetimes. He argues that modelling how a DNA marker is affected by cancer is too complex today, but that as models improve and computers grow more capable, they are going to solve it. He also predicts widespread humanoid robots within five years.
Haas warns that a global chip shortage is slowing AI because data centres need vast numbers of processors. He calls the market supply‑constrained and notes plans for multi‑gigawatt data centres in France and the US, and even proposals to put them in space. Arm’s power‑efficient designs are used in phones, cars and gadgets, and the firm now sells some of its own chips. Meta asked Arm to develop the Arm AGI chip, and demand since March has topped $2bn.
Haas used to sit on AstraZeneca’s board but stepped down in April. He is sceptical about manufacturing chips in the UK, saying fabs are costly, need specialised workers and resources, and the ecosystem lies elsewhere. He adds that forecasts of mass job losses from AI are overstated, and that new roles would grow as robots and tools spread.
Comprehension
- Who says AI could help cure cancer in our lifetimes?
- What factor does Haas say is slowing AI growth?
- Which company asked Arm to develop the Arm AGI chip?
- Which countries did Haas cite for planned multi‑gigawatt data centres?
- What is Haas’s view on manufacturing chips in the UK?
Grammar focus
Used to / would
Both describe past habits. Only 'used to' can describe a past state — a situation rather than an action. 'The building used to be a cinema' is correct; 'would be' is not. (Example 2 constructed.)
- Haas **used to sit** on AstraZeneca’s board but stepped down in April.
- Engineers **would visit** factories regularly before automation spread.
Grammar exercise
Complete each sentence using 'used to' or 'would'.
- The mayor (visit) ___ the market every Friday.
- The building (be) ___ a small cinema before the war.
- Neighbours (share) ___ ladders and tools freely.
- Workers (walk) ___ to the docks before dawn.
- The town (have) ___ three daily newspapers.
Discussion
- Will AI cure diseases faster with better data or bigger computers?
- Should the UK invest in fabs despite high costs and resource needs?
- Would humanoid robots in services make life better or worse? Why?
- How could chip shortages affect everyday products you use?
Vocabulary, reading, and exercises with instant feedback