Cancer services are being told they can’t hire doctors while patients face long waits and doctors are left unemployed. Doctors who want to diagnose and treat cancer are being stopped at multiple stages from helping patients. The Government caps how many doctors are allowed to train as specialists like radiologists and oncologists (specialists who read and diagnose from scans and treat cancer). Thousands apply each year, and many are rejected because there simply aren’t enough training jobs. Last year, 4,011 doctors applied for clinical radiology training. Only 356 were accepted. For the few who do manage to specialise, it doesn’t get easier. In 2025, half of the UK’s specialist cancer centres were banned from recruiting oncologists, and over a third of radiology departments were blocked from hiring radiologists due to recruitment freezes, despite rising demand and high waiting lists. So we now have an under-doctored country where: - Doctors are turned away from specialist training - Qualified cancer doctors are prevented from working due to recruitment freezes - Patients wait longer for diagnosis and treatment This is not an accident. It’s the result of government caps and cost cutting as well as poor workforce planning. And the jobs crisis is one of the reasons why resident doctors have recently voted to extend their strike mandate. This is not a workforce mystery. It is the predictable result of government caps and NHS cost cutting.

This paper from Harvard and MIT quietly answers the most important AI question nobody benchmarks properly: Can LLMs actually discover science, or are they just good at talking about it? The paper is called “Evaluating Large Language Models in Scientific Discovery”, and instead of asking models trivia questions, it tests something much harder: Can models form hypotheses, design experiments, interpret results, and update beliefs like real scientists? Here’s what the authors did differently 👇 • They evaluate LLMs across the full discovery loop hypothesis → experiment → observation → revision • Tasks span biology, chemistry, and physics, not toy puzzles • Models must work with incomplete data, noisy results, and false leads • Success is measured by scientific progress, not fluency or confidence What they found is sobering. LLMs are decent at suggesting hypotheses, but brittle at everything that follows. ✓ They overfit to surface patterns ✓ They struggle to abandon bad hypotheses even when evidence contradicts them ✓ They confuse correlation for causation ✓ They hallucinate explanations when experiments fail ✓ They optimize for plausibility, not truth Most striking result: `High benchmark scores do not correlate with scientific discovery ability.` Some top models that dominate standard reasoning tests completely fail when forced to run iterative experiments and update theories. Why this matters: Real science is not one-shot reasoning. It’s feedback, failure, revision, and restraint. LLMs today: • Talk like scientists • Write like scientists • But don’t think like scientists yet The paper’s core takeaway: Scientific intelligence is not language intelligence. It requires memory, hypothesis tracking, causal reasoning, and the ability to say “I was wrong.” Until models can reliably do that, claims about “AI scientists” are mostly premature. This paper doesn’t hype AI. It defines the gap we still need to close. And that’s exactly why it’s important.

The more time students spend on screens, the less they learn. Ed tech does not belong in schools (until it is thoroughly tested & proven to help). Excerpt from Jared Cooney Horvath's excellent new book, The Digital Delusion, in @TheFP thefp.com/p/we-gave-students…

The most dangerous addiction today isn't a substance. Research on 100,000 people confirms that heavy short-form video use is just voluntary cognitive decline. We are actively training our brains to fail at hard tasks. If you can simply sit with a problem for 10 minutes without swiping, you have a massive competitive advantage. Basically, boredom is the new IQ.

Here's a dark but fun essay I published in @TheFP last week. We are reposting it on AfterBabel.com. It uses ChatGPT role-playing the Devil to lay out how to "destroy the next generation." Turns out we're doing a really good job of it already afterbabel.com/p/the-devils-…

164,800 people waited over 12 hours in A&E last month. At the same time, 14 doctors are fighting for one A&E job. Patients are stuck in corridors while NHS doctors who want to work are blocked by Government caps on training jobs. The system is failing both staff & the public.

Over 6,000 farm business have shut since Labour came into power. And many of them because of Rachel Reeves & the Labour government lying about the “£22bn black hole” - used to justify the inheritance tax on family farms. Thousands of farm businesses shut because of a lie.

BREAKING The OBR says it informed Rachel Reeves as far back as ***September 17*** that the downgrade in productivity forecasts was offset by 'increases in real wages and inflation'. The deficit was in fact just £2.5billion By October 31 that deficit had turned into net positive of £4.2billion. That basic forecast did not change from that point So from what the OBR is saying it looks like Rachel Reeves and the Treasury were briefing ahead of the Budget that there was a £20billion black hole in the public finances that didn't actually exist The £30billion worth of tax rises in the Budget are predominantly a consequence of her decisions to increase public spending, particularly on welfare, and have £21.7billion worth of headroom As @Peston @PippaCrerar @hzeffman have all pointed out, it makes the Budget build up - and the narrative that big tax rises were coming because of a deterioration in the public finances - look frankly surreal in hindsight

The recruitment system in England is a utter disgrace. We now have fully trained NHS doctors who cannot find work. Every medical leader who stood by and let this happen should be ashamed. This is a total failure of workforce planning. Doctors deserve better than this.

Last time I checked, juries were turning in fair and sensible verdicts while judges were increasingly appearing to bow to political and ideological biases. Is that why @DavidLammy wants to abolish the juries rather than reform the judges? Truly terrifying.

There is more and more evidence that putting computers and tablets on students' desks (1:1 devices) was a terrible mistake. I agree with @AdamMGrant that "it's time to remove laptops [and tablets] from classrooms."

NEW REPORT: History should be taught to all children up to the age of 16. The call has been endorsed by the distinguished historians Niall Ferguson @nfergus , Lord (Andrew) Roberts @aroberts_andrew and Dominic Sandbrook ('The Rest is History') @dcsandbrook . At present, most children (55%) stop learning history at 14. That is why our exclusive new poll reveals such little knowledge of history among young adults. Three quarters say they don't know about Admiral Nelson. More than half say they don't know about Oliver Cromwell, a vital figure in the development of our democracy. Britain is an outlier among OECD countries in not teaching history up to the age of 16. This should be changed so we give a rounded education to the citizens and voters of the future. Press release and report below. historyoftotalitarianism.com…

School-based Social and Emotional Learning (SEL) intervention found ***zero*** impact on children’s mental health. psycnet.apa.org/fulltext/202…

Thousands of NHS doctors are being blocked from becoming Consultants/GPs due to Government caps on speciality training places. Patients wait months for care while qualified doctors are left jobless. Wes Streeting knows this - so why isn’t he fixing it? telegraph.co.uk/opinion/2025…

New evidence on the effect of cell phone bans in schools just dropped. "Enforcement of cellphone bans in schools led to a significant increase in student suspensions in the short-term ... but disciplinary actions began to dissipate after the first year, potentially suggesting a new steady state after an initial adjustment period." "We find significant improvements in student test scores in the second year of the ban after that initial adjustment period."

One of the many delightful surprises coming in from newly phone-free schools: Kids check out many more books from the library!

You’ve waited months for your operation. You’ve waited hours in A&E. And yet; thousands of doctors who want to become surgeons and emergency medics are being turned away. In 2025, Emergency Medicine had 5,081 applicants for only 357 posts; around 14 applicants for one post. 5,399 doctors applied for just 630 Core Surgical Training posts; a ratio of 8.57 to 1. These are doctors who’ve completed years of study and clinical training, passed rigorous exams, and are ready to serve patients. But the Government refuses to fund enough speciality training places for them. So while patients lie on trolleys in corridors and operations are cancelled, fully trained doctors are left without a pathway to specialise. This is a catastrophic failure of workforce planning. A deliberate Government imposed cap on doctor speciality training numbers that wastes talent, demoralises staff, and prolongs NHS waiting times. If the Government is serious about tackling long waiting lists and rebuilding the NHS, it must expand funded training posts and use the workforce we already have.

As Susanna Reid asks, "how have we got to a situation where we are training up doctors, investing a huge amount in them, we have an NHs in crisis... it is hard to get a GP appt, and we can't give trained doctors jobs?" #GMB