Another hilarious moment with Claude.ai. I asked it to criticise a piece I'd written and a) it thought the bits I wrote myself (albeit while being somewhat distracted) were "the most AI-sounding"; and b) the bits I'd asked ChatGPT to draft were "the strongest writing". Some rather brutal feedback!

Great reporting by FT's Tej Parikh on a great new report by McKinsey Global Institute. One of the reasons investment in the UK (and other countries) is so low is that it's just really expensive. #dataisbeautiful giftarticle.ft.com/giftarticle/...

The AI adoption gap has its roots in capabilities built—or not built—well before the latest AI wave. Pre-existing management practices, organisational and human capital, and data and software assets, all help explain which businesses have adopted AI. substack.com/@teraallas/n... #dataisbeautiful

UK #AI adoption remains surprisingly low (see updated chart with June 2026 data). Read about the barriers, the productivity paradox, and the 'ladder of leakage', in our recent article: mck.co/4b8w4wh #dataisbeautiful

Interesting data on the UK's infrastructure investment and capital stocks. Higher than I had expected - but highly concentrated in electricity. #dataisbeautiful bit.ly/44FgHHH

Really pleased to see this published. The UK’s underinvestment puzzle needs solving if we want to build a prosperous and resilient future.

Why do Brits fail to invest for the future (incl. climate adaptation)? This international comparison is striking and, while descriptive rather than causal, suggests more short-termism than comparator countries. Full paper here: bit.ly/4atmsvu #dataisbeautiful

🎙️ New Productivity Puzzles episode! Why do UK firms not invest? @bartvanark.bsky.social talks with @teraallas.bsky.social, Stephen Roper & Catherine Mann about why firms hesitate to invest and how management decisions shape productivity. Listen: www.productivity.ac.uk/podcast/why-...

The increased frequency of hot weather events has been predictable for decades. Why do Brits find it so hard to invest for the future (in this case, climate adaptation)? #dataisbeautiful

Is AI stealing jobs? Is it boosting productivity? Those are the two biggest economic questions about AI. Remarkably, the UK's official statistics are struggling to answer either of them. A look at what the data can — and can't — tell us: open.substack.com/pub/teraalla... #dataisbeautiful

Oh dear. Slightly depressing answer from Claude.ai on (the hugely important question of) how UK investment returns compare to other countries’. I wouldn’t be asking the question if we had already proved it in our previous paper…

The UK has ranked near the bottom of the investment league table at least since the 1950s, with serious consequences for our productivity and prosperity. Could this reflect cultural attitudes and dispositions among UK decision makers? Yes! Read our research here: bit.ly/4atmsvu #dataisbeautiful

Only 11% of UK businesses intend to increase investment in the near term. Yet, few firms cite access to finance or inadequate returns as a reason. Read our @productivity.bsky.social research on how UK managers' attitudes may be constraining profitable investment: bit.ly/4atmsvu #dataisbeautiful

How should governments respond when AI’s economic impacts remain deeply uncertain? A new piece on “strategy under uncertainty”: no-regrets moves, optionality, and why policy should avoid overcommitting to any single AI future: teraallas.substack.com/p/ai-and-the... #dataisbeautiful

What is holding back the #investment that we so desperately need for #productivity growth? One reason could be that the UK population, and its managers, are consistently more risk averse than comparator countries'. Read our new research: bit.ly/4atmsvu

New UK job-ad data suggests the labour market for young graduates is weakening significantly. But the story is more complicated than “AI is replacing graduates”. #dataisbeautiful open.substack.com/pub/teraalla...

Enjoyed contributing to this balanced piece by Sam Fleming, Valentina Romei and Philip Stafford in the FT (link below). We might be entering another cycle of differential inflation for households. Latest ONS data on this published today (see chart). giftarticle.ft.com/giftarticle/...

New UK job-ad data suggests AI exposure is not affecting all jobs equally. AI seems to be displacing routine office and service work—not graduate occupations. So far, this looks like an acceleration of a much older labour-market trend. #dataisbeautiful open.substack.com/pub/teraalla...

It was great to contribute to this article on UK #productivity by Ben Zaranko, alongside Gregory Thwaites, Peter Dixon and others. The retail and wholesale story is consistent with capital deepening; does not apply to all other sectors... observer.co.uk/news/busines...

Really enjoyed contributing to this excellent article on UK #productivity!

The UK's investment problem is well known (Y-axis). What surprised me was the X-axis. We complain about the "nanny state", but when global energy or food prices rise, the conversation quickly shifts to what the government should do. #dataisbeautiful

Some potentially good news on UK #productivity turning a corner. However, re-basing the data and looking at the longer term illustrates how tentative this is. #dataisbeautiful

Demand for AI skills is rising rapidly. But so too is demand for business analysis, process optimisation, people management and leadership. Judgement, relationships, and trust are the human traits with a premium in an AI-driven world. teraallas.substack.com/p/ai-and-ski... #dataisbeautiful

I've been doing some novel research on the UK's under-investment problem and came across this. I've been saying we have had the problem for 'decades', but the timeline here (1960-1992) indicates just how long low investment has been an issue! #dataisbeautiful

Rethinking Productivity in the Public Sector & the NHS. Join us on 9 June at @alliancembs.bsky.social to explore how productivity can be improved in the NHS, & what this means for other large, complex organisations. Book on Eventbrite: www.eventbrite.co.uk/e/improving-... @teraallas.bsky.social

Historically, in many sectors, higher productivity growth has been associated with more—not fewer—hours worked. A short note using UK sectoral data on why productivity growth does not automatically translate into job destruction. #dataisbeautiful open.substack.com/pub/teraalla...

What’s the point of debating #AI, if it doesn’t have any impact on people's happiness? A striking result—and what it might be missing. teraallas.substack.com/p/ai-and-soc... #dataisbeautiful

If AI is so powerful, why isn’t it being used more? Because adoption depends on trust, not just capability (or cost-benefit). #AI #productivity #dataisbeautiful teraallas.substack.com/p/ai-adoptio...

The NHS productivity challenge | @escoeorg.bsky.social Join us on 30 April in London for a panel on why NHS productivity has yet to recover & what reforms are needed for a more sustainable NHS. A fantastic panel, with @teraallas.bsky.social. Register: www.eventbrite.co.uk/e/what-does-...

Fun chart from the ever-excellent @resolutionfoundation.org @ruthcurtice.bsky.social Top Of The Charts. Next: best arbitrage opportunities, if any, factoring in travel costs? #dataisbeautiful [Link to RF story: www.resolutionfoundation.org/comment/how-...

UK #consumer #confidence has taken a real hit across the board since the Iran conflict started - and that's against an economic backdrop that's arguably more challenging for many households than 2022. #dataisbeautiful

Fantastic appointment! @bartvanark.bsky.social leaves an impressive legacy and Jun Du has all the right qualities to build on it for continued success.

UK #GDP and #inflation forecasts have deteriorated considerably in the last month. Analysts are still not expecting a recession, but moderate #stagflation seems to be on the cards. Tough call for #BoE. #dataisbeautiful

If AI is boosting productivity, why isn’t it showing up in the macro data? A short note on where the gains get lost (aka "the ladder of leakage"): open.substack.com/pub/teraalla... #dataisbeautiful

#AI may be boosting firm-level #productivity, but at the moment, UK employment is declining in sectors with the least AI exposure. See also here: open.substack.com/pub/teraalla... #dataisbeautiful

AI exposure is often read as a measure of job risk. But in principle, it tells us what AI could do—not what will happen to jobs or wages. Outcomes depend on how jobs are structured and how firms adopt the technology (and other factors). substack.com/home/post/p-...

🎥 CATCH-UP: What can we learn about NHS productivity from developments across the wider economy? With @teraallas.bsky.social​​ from our #NHSProductivityCommission, this webinar heard from a panel on what we can learn from sectors and industries beyond health care. Watch now⬇️ https://bit.ly/4b3Ao05

AI exposure ≠ impact. In the recent "Forecasting the Economic Effects of AI" paper there is no clear relationship between AI exposure and expected job growth. A useful reminder that task-level measures don’t translate directly into labour market outcomes. teraallas.substack.com/p/ai-and-job...

Of the businesses that say they have adopted, or are planning to adopt, #AI, very few say they have experienced or are expecting headcount reductions as a result. Consistent with most AI benefits coming through in top line sales and innovation, so far. #dataisbeautiful

Here's the most basic reason #AI is not yet boosting productivity (or displacing jobs): the majority of businesses are not actively using it! Technology diffusion is slow (until it isn't). More of what to expect our article: mck.co/4b8w4wh #dataisbeautiful

UK business investment was down sharply (-2.5%) in Q4 2025, against a volatile trend. What keeps UK investment rates at close to the bottom of the league table? My future research will illuminate, but risk aversion could be a key driver. #dataisbeautiful

Interesting how the oil forward curve has shifted -- across all maturities, but much less at the far end. Of course, the future remains highly uncertain and, due to lags in price pass through, we'll likely feel the effects for a while to come. #dataisbeautiful

Not headline-grabbing, but important and predictable: the jobs which with the biggest impact on job demand from #AI are those that were already in decline due to previous digital technologies: admin and customer service roles that don't require much education. #dataisbeautiful

There's a lot of confusion about what impacts #AI will have on jobs and wages. No surprise, given that it's not a simple question! There are many, many determinants beyond the now-popular 'exposure' measures. I like this framework. #dataisbeautiful

Asking ChatGPT to critique Claude's output, and the other way, is fun. The revised draft is typically still not brilliant, but it is often a step forward. But I wanted to go up a level and ask what to explicitly tell them to avoid. This rings so true; and I'm impressed with the self-awareness!

To contribute to the debate about whether #AI will displace a lot of jobs (rather than just tasks), I’m resharing something I wrote in 2017. The core point still holds: automation ≠ job losses. This isn’t the whole story — but it’s a key part of it. bit.ly/4dQvwwO #dataisbeautiful

New Productivity Puzzles episode: What does it take to improve NHS productivity? Featuring @healthfoundation.bsky.social's report: From diagnosis to delivery: www.health.org.uk/reports-and-... As @teraallas.bsky.social notes, the NHS is undermanaged. Listen to the ep.: podfollow.com/productivity...

#Finland tops the World Happiness Report rankings — again. One intriguing clue as to why: Finns feel little pressure to show how good they are or seek others' admiration. Freedom from that pressure may be central to a good life. How very Buddhist! #WHR #wellbeing #dataisbeautiful

Great to see The Economist reporting on the World Happiness Report. I wonder if Patrick has ever met a Finn or been to Finland, though? Article here: bit.ly/3NNu5EZ See also my earlier post: bit.ly/4rK6i6A #WHR #wellbeing

Fascinating to see the latest World Happiness Report (lnkd.in/gZAgdzuU ). #Finland continues to top the leaderboard of the world's "happiest" countries. It's certainly not because of the weather - the Finns get, on average, a shockingly low 5 hours of sunshine a day. #wellbeing #dataisbeautiful