Expert Career Opportunities for Exceptional Professionals

Episode 2 of 'Build something you find useful then share it'. Introducing PlateCheck - an app for people with food intolerances to find things they can actually order from restaurant menus. I've struggled with gut issues for years. For a long time my strategy was 'ignore it and deal with the consequences' - self-medicating with Trebor Extra Strong Mints and spending hours feeling nauseous and bloated. When that stopped being an option, I got proper help. A year of restrictive diets later - anyone who knows what FODMAP means will know the pain - I was diagnosed with SIBO and lactose intolerance. Both annoyingly simple to fix. Oh how I wish I'd known years earlier. Going to nice restaurants is one of my real luxuries in life. But even now, knowing what to avoid, it's easy to get it wrong especially when not every ingredient is listed. On Saturday I ordered something at a family lunch, confident it was fine, and spent the rest of the day feeling awful. So on Sunday I built PlateCheck. Open the app, select your intolerances, upload a photo of the menu, and it tells you what's safe to order, what to ask about, and what to avoid. Right at the table. Or before you even get to the restaurant, if you daydream about good meals like I do. Take a look: https://lnkd.in/eJuY5MFD. Let me know if it helps next time you go out to eat, and share with your food intolerant friends. (Important disclaimer: this is not for allergies and may still make mistakes! Always check if you’re unsure.)

Before we launched anything, we spent months learning from accomplished, mid-career professionals what they actually needed. Some of what I learned surprised me. I was designing a program for people who want to redirect their careers toward work on some of the biggest challenges of our time like AI safety, pandemic preparedness, and factory farming. My first instinct was to build something comprehensive. Weeks long, lots of content; we could even lift pre-made fantastic resources from impact-focused career advisors. But then I actually started speaking to accomplished professionals about what they actually needed. Interviews with accomplished professionals who wanted to make this kind of transition. Surveys. Mockups of programs that didn't exist yet, put in front of people to see which ones they'd actually sign up for. The message I heard: ‘I don't need comprehensive career planning. I already know what I’m good at. I need something to help me translate my skills into work that actually improves the world, and get clarity on my next steps.’ The product we ended up building wasn't on our original brainstorm list. And it was shorter than anything I would have designed on my own; only four days! I didn't trust it at first. So I tested it. A pilot with 5 people, a modified version with 35. Pass/fail criteria set before each round so I couldn't rationalise mediocre results after the fact. In the first public launch we got 700 applications in a month, with no paid marketing. The short format worked because it helped solve the actual problem. Many people thinking about this kind of pivot aren't stuck because they lack information. They're stuck because there's no deadline, no structure, no rapid feedback. A year goes by and they're basically in the same place. Four days removes the room to postpone, procrastinate, or ‘perfect’ that application until it's too late. With 8 cohorts and 150+ participants, we’ve iterated to create a bootcamp I’m really excited about. It’s now rated 8.9/10 by our most recent 4-day cohort. Free, selective (~25% acceptance rate), cohorts running through April–June. Apply before the deadline >> https://lnkd.in/eDPWNx9R

Most of my career pivots began with work experience. I’d offer my skills for free, or on a short-term basis, to learn the work and prove my value. It always led to a job, one way or another. If you’re starting out or wondering how to pivot into something more meaningful, this is still the approach I’d recommend. ☝️ Orgs working on global health, AI safety, animal welfare and more are looking for interns, fellows, and volunteers. Worth checking out if you want a foot in the door.

Does money buy happiness? Yes... but less than you would expect! Doubling your income from £50,000 to £100,000 is linked to a happiness increase of under 2 points on a 100-point scale. That's true wherever you start: a doubling of income is associated with a gain of less than 2 points. £20,000 to £40,000? Less than 2 points. £200,000 to £400,000? Less than 2 points. The finding comes from a 2023 collaboration between Matt Killingsworth and Daniel Kahneman. The two had reached conflicting conclusions in earlier work, so they pooled their data to settle it. This and similar studies got me thinking about what it means for how I, and other members of the Effective Altruism and Giving What We Can movements give. Here are my quick takes: 1. That 2-point figure already reflects how high earners live: longer hours and less free time included. The actual return on all that grind feels pretty small to me. Perhaps a better way to build personal happiness is to use precious resources like your talent and time elsewhere where the happiness returns are higher? A more meaningful job (Rutger Bregman discusses this in detail in his latest book: "Moral Ambition") or simply spending more time with those you love and earning less. 2. Doubling a person's income from £500 to £1,000 is associated with the same happiness gain as doubling yours from £50,000 to £100,000. The only difference is the price. That same £50,000 could instead deliver a comparable gain to around 100 people living below the international poverty line. GiveDirectly sends cash straight to people in extreme poverty, which is about the most obvious way I can think of to make that trade. 3. Giving tends to make the giver happier. In one experiment, people who were handed money and told at random to spend it on others reported more happiness than those who spent it on themselves. Brain imaging shows donating activates the same reward centres as receiving. Put together I’ve come to this conclusion: the richest person isn’t the one who has the most money. It’s the person who wants for the least in their life. Want to learn more about buying happiness? Read our blog on this topic. Link in the comments.

You need user buy-in to scale your impact! For many impact-focused programs, it’s not enough to have a plausible theory of change; you also need buy-in from your users. All products/interventions have users, even if they’re not your intended beneficiaries. Successfully meeting real user needs is usually a necessary (but not sufficient) condition for scaling a cost-effective program; users who are bought-in are more likely to follow through with the recommended high-impact actions, and programs with strong demand can scale more easily. A useful way to think about this is that you want to seek “product-market-impact fit”. This means building programs that users genuinely want to engage with and that create cost-effective impact. The phrase is stolen from Peter McIntyre, based on the startup concept of product-market fit. It’s a process I’ve been learning, implementing, and refining through my ~7 year career in impact-focused talent search and community building, as co-founder of Animal Advocacy Careers → Managing Director at Leaf → Courses Project Lead at the Centre for Effective Altruism. But it’s not just important for community building. For example, some researchers produce rigorous work that nobody acts on, because they never asked decision-makers what they needed. Others orient their research around the needs of funders, so end up getting more work directly commissioned and influencing tens of millions in funding decisions. More details in my EA Forum post (link in comments)

Why are some teams able to consistently outperform others? I don’t think it’s always because they work harder, or hire better, or care more. In fact, I’ve seen some of the most conscientious, talented and hardworking teams come up short or bang their head against the same wall time and again. I think it’s because they've figured out how to be strong across all the things that matter all at once, and that’s hard. And it’s even harder to sustain it over time. I’ve spent most of the last decade trying to answer that question. Where I’ve landed is what I call the Six Pillars of High Performance, the six critical elements every high performing team focuses on and strengthens to achieve their potential. Most teams are genuinely strong in at least a couple of them at any given time. But it’s the others, the elements that can boost or hinder teams, that stand in the way. I built a diagnostic tool to help you figure out where you are right now and where you need to go. It takes less than ten minutes, and you’ll get a full report if you want to dig in. Try it here: https://lnkd.in/ezkh8etT. Let me know what you think.

I had my proudest-ever moment as a philosopher 11 days ago: I learnt that the novel theory of moral uncertainty I'd argued for is now the 'default view' of a major research team and should help $100m+ of philanthropic funding get spent better. Wait, what am I talking about? At Effective Altruism Global London in May, Rethink Priorities launched their new 'cross-cause fund' 👇. The idea is to work out where the marginal dollar will do the most good, across all the cause areas they've investigated. This is a very complicated task with many inputs. One crucial input into this is your theory of moral uncertainty. Now, we're all familiar with empirical uncertainty - not knowing the facts. However, we are also uncertain about ethics. How much should we value animal vs human welfare? How do we weigh lives that exist today against future possible lives? And so on. Clearly, our ethics will change what we ought to do, but we all have doubts. This raises the question: how should we act given our moral uncertainty? You can think of it like there's a parliament in your head made up of ethical theories who disagree, and you're wondering what choices to make overall. Moral uncertainty is a very new field in philosophy; ~all the work is from the last 20 years. As you'd expect, there are various candidate theories on the table. I'll mention one here: the Moral Marketplace Theory (MMT). On MMT, the basic idea is you split your resources according to your credences. So, imagine you're torn between two views, where (1) is that helping animals in factory farms is the top priority, and (2) is helping people alive today. Suppose you're 20% confident in (1), and 80% in (2). The theory recommends you split the pot: 20% goes to the animals, 80% to the humans. This is the method Rethink Priorities adopted as their default approach. It's based on the theory Harry Lloyd, Patrick Kaczmarek and I argued for in a 2024 paper in Synthese after we spotted a gap in the literature 👇. It's not the leading theory. We don't pretend we've said the final word. No one's even replied to our paper yet. Hence, I was stunned to learn that Rethink Priorities had looked at our little idea and thought "yup, that seems right. We'll develop and use that". Why did Rethink opt for this over the alternatives? I don't have space to explain here, but I'll link to Marcus Davis's explanation 👇 (for the nerds: the advantages of MMT are that it avoids intertheoretic comparisons, fanaticism, and justifies diversification). The dream for a philosopher is that you come up with a new theory, other people agree with it, they actually use it, and this makes a difference. I never thought this would happen to me, and I am chuffed. 😁 Well done to Rethink on their work so far. I look forward to seeing the cross-cause tool progress - especially for Happier Lives Institute's charities to be included! *Pic of me sharing my chuff-ness w/ Pericles in the British Museum

One of the perks of my job is the people I get to meet along the way. Michael Thatcher🔸 is one of those rare combinations of head and heart; kind, down to earth, and incredibly thoughtful about impact. It's this advice he shares with anyone thinking about doing good in the world — follow your heart, use your head, and then go make a difference.

Should we prioritize making existing people happier, or creating more happy people who will live good lives? Is lying ever okay? Answering questions like these often means choosing between competing values, resulting in moral uncertainty — uncertainty about which ethical principles are correct, or how to weigh them when they conflict. Here are three strategies for navigating moral uncertainty and making decisions aligned with your values: 1️⃣ My Favorite Theory (MFT): MFT suggests that you pick the moral theory you find most credible and follow its guidance. While fast and intuitive, MFT does not account for other plausible theories, so is often criticized for being oversimplistic. Read a defense of MFT here: https://lnkd.in/gDCHvTeR 2️⃣ Expected Choice-Worthiness: Assign probabilities to each moral theory you find plausible (like utilitarianism or virtue ethics) and calculate which actions have the highest ‘expected moral value’ across all theories, similar to an expected value calculation. While slower than MFT, this systematic process enables you to incorporate several moral perspectives into your decision-making. Learn more here: https://lnkd.in/gzivhyzr 3️⃣ Moral Parliament: This approach involves simulating your own “Moral Parliament,” giving each of your moral views a different amount of voting power based on how confident you are in them. After rounds of 'negotiations,' you can tally votes to arrive at decisions. Try it for yourself with Rethink Priorities's interactive tool: https://lnkd.in/giuH4rkv. Or read more about moral parliaments in Toby Newberry and Toby Ord's paper: https://lnkd.in/gRrwC37D Moral uncertainty matters if we want to do the most good. It's impossible to have perfect confidence about where our actions matter most, but frameworks like these help us make balanced decisions. For more, check out the EA Forum: https://lnkd.in/g3Qx2ST3

Anthropic just announced Mythos Preview — a model so capable at finding and exploiting software vulnerabilities that they've refused to release it publicly. It found zero-days in every major operating system and every major web browser, some decades old. They're restricting access to a handful of companies for defensive cybersecurity work only. Models this powerful raise an obvious question: how do you actually make them behave well? One of the approaches I find most interesting is constitutional AI. The basic idea: instead of relying purely on human feedback to train models (expensive, inconsistent, hard to scale), you give the model a written set of principles — a constitution — that it can use to evaluate and improve its own outputs. This makes alignment more legible. You can read the document, argue with it, point to where it gets things wrong. Anthropic published Claude's constitution earlier this year. Think of it as the US Constitution and the Federalist Papers combined into one document — it doesn't just set out the rules, it explains and defends the reasoning behind them, so the model can generalise rather than mechanically follow instructions. But because it's written with Claude as its primary audience, it's a pretty weird experience reading it. It's optimised for precision over accessibility, and it covers all sorts of topics. It establishes four priorities in explicit rank order: be safe, be ethical, follow Anthropic's guidelines, be helpful — in that order. It defines a hierarchy of 'principals' (Anthropic, then API operators, then end users) and works through how Claude should reason when their interests conflict. It addresses honesty, moral uncertainty, when to refuse requests, and even questions about Claude's own nature and possible moral status. The lead author is Amanda Askell, who heads Anthropic's personality alignment team; the New Yorker described her as supervising Claude's 'soul'. She grew up in Scotland, studied philosophy at the University of Dundee, did an Oxford BPhil, then a PhD at NYU on infinite ethics. She was married to Will MacAskill — one of the founders of effective altruism — and remains a Giving What We Can pledge member. Two resources if you want to go deeper: Anthropic's blog post explaining the constitution and the thinking behind it: https://lnkd.in/efkERXfH And the latest episode of Justified Posteriors — 'The Most Important Philosophical Treatise of the 21st Century?' — where economists Seth Benzell and Andrey Fradkin do a proper critical reading of the constitution. Good podcast generally if you're interested in the economics of AI. https://lnkd.in/eqnRYNST