I recently went on Chris Bryant's YouTube stream to talk about using AI for animal advocacy

Me and the Bryant team have been using Cowork for two weeks now, and it's pretty clear that this is the future of work. We now have multiple instances where the team has managed to complete one or two weeks' worth of work in a single day. However, it also has a few issues and rough edges. Lots of things don't work the way you think they would work.

When Gemini or Claude doesn't work, the true AI Pros build custom workflows....or do they? This could be automations in Make.com or n8n, or it could be "skills" in Claude Code. It might involve jargon MCPs, or the sinister-sounding OpenClaw. It could be Python scripts or cron jobs. Btw, if that last paragraph made zero sense to you, don't worry! This post is still for you.

I divide my week between being the director of research at a social science research consultancy and being a data scientist at a climate NGO. Over the past three months I have pivoted to using Claude code and OpenAI's Codex CLI for nearly all of my research work. I have honestly never been this productive, and I think that eventually all scientists are going to be using these tools as their primary way of doing research.

I have had my Plaud AI notepin for about three months now and I absolutely love it. I can record my thoughts or ideas with the press of a button. It makes taking notes on all sorts of things in my life totally frictionless. I'm going to be writing a blog post on how I use it in a few weeks, so stay tuned!

In 2025, LLMs got terrifyingly good at software engineering. So why are they so unimpressive at data science? Data science doesn't really use tests that much LLMs are so powerful for software engineering is because of a 1-2 punch: As long as software tests are good, and code passes those tests, you can be confident that LLM code is good. You can have LLMs quickly produce vast amounts of code, and you don't have to review it line by line.

Here's how you can automate pulling your latest Oura Ring data into a Google Sheet every day To do this we are going to use something that most people don't know about but is insanely powerful: Google Apps Script. Google Apps Script is a way that you can create new features for Google Docs, Slides, and Sheets by writing JavaScript.

What if we fundamentally recreated exams from the ground up, with AI at the core? I see the potential to make a radically better way to do school exams. What's more, most of the technology already exists. Before we begin: I'm going to use the term "LLM" a lot. That stands for "large language model". Examples include ChatGPT, Gemini and Microsoft Copilot.

The consensus is that overall, carbon offsetting hasn't worked at reducing or negating emissions. Does that mean offsetting meat consumption is a dead-end? I argue no. If we look at the 2 main reasons why carbon offsetting failed, we can see that meat offsetting does not have these problems. As a result, meat offsetting may be an effective strategy to reduce animal suffering.

I was recently on "The Vegan Report" podcast talking about AI and the animal movement. We discuss AI tips, how advocates are using AI in their activism, fears and concerns people have over AI, and even AI consciousness!

AI can do much of the work that used to be reserved for interns and juniors. For cash-strapped, burnt-out workers in the NGO space, AI assistants are a tempting alternative to managing early career employees. But how can we reap these benefits without shutting out the world changers of the future? Will AI automate all of our jobs? Or will it create tons of new ones?

Most people write gratitude journals about the good things in their life. I've done this, and while it feels good at first, it gets a bit repetitive. It cultivates a useful mindset, which is to just regularly be grateful for things in your life. However, I think we can cultivate a second, different mindset; A mindset of being grateful even for shit things.

If you're an AI power user, this post isn't for you. It's for the large number of people who see the hundreds of AI tools out there and feel totally overwhelmed. I know this is how people feel, because I just taught a 12-week AI course and this was common feedback. So I thought I would cut through the noise and give you 5 things to try.

This summer was a busy one! I gave 3 talks on a topic I have accidentally became an expert on over the last 2 years: Food security! How factory farming actually undermines food security, and how alternative proteins and other plant-based foods are the true answer. CARE, Warsaw This summer I had the pleasure of speaking at the 2025 Conference on Animal Rights in Europe (CARE)

In 2023, I gave a talk on AI and art. In 2023! Feels like a lifetime ago. It was about how AI could affect the market for art and jobs for artists. Yet in the discussions after the talk, the question that generated the most discussion was: "Is AI art real art?" To me, the answer was a clear "yes".

Recently, Animal Charity Evaluators interviewed me about how I use AI in my work, and how I think AI will affect the animal advocacy movement They also interviewed smarter people than me: Aaron Boddy, Shrimp Welfare Project Hannah McKay, Rethink Priorities Janire Castellano Bueno, Wild Animal Initiative Sagar Shah, Rethink Priorities We all independently came to the same conclusion: if the animal movement doesn't get serious about using AI for animal liberation, we risk being left behind! You can find the interview here:

People often ask me if AI can help them with data analysis and making graphs. To be honest, this isn't something that I have a huge amount of experience in, because I'm a data scientist who makes all of his own graphs. BUT, I have been reading some very interesting pieces by other people who do know about this kind of thing, so I thought I would collect together my 3 favourite, recent blog posts on Using LLMs for charts and graphs, and summarise the common themes.

Myself and 3 other AI consultants are teaching a free 12-week practical course in generative AI. You'll go from beginner to confident user. It's open to anyone working on fixing our food system: promoting plant-forward diets, increasing animal welfare, fighting factory farming or doing right by farmers. It's going well so far: We've had 1000 sign ups! We're actually about to start week 4, on how to spot and reduce AI hallucinations.

AI is transforming everything. How will it transform animal advocacy? Is it going to supercharge our campaigns, or render some of them totally obsolete? Will it help or hurt factory farming? These seem like pretty big abstract questions, but given the pace of AI development, we need to be talking about them. And increasingly, people are! Mostly on the Effective Altruism forum…

Scroll through any thread about AI and you’ll see two loud camps: the cheerleaders and the haters. One side insists AI is pure excitement: innovation, breakthroughs, the future is here! The other insists it’s pure outrage: It's all hype, it ripped off artists, it's causing the climate crisis. Both act like their reaction is the only legitimate one, as if you have to pick a team: thrilled or furious.

"Do I need to learn about neural networks to understand ChatGPT?" I hear this all the time, and have been asked this a few times. I also see people making videos, courses and blog posts breaking down the technical details of how large language models (LLMs) work. People assume that by understanding the technical and theoretical concepts behind these models, they'll gain secret insight into how to use them better.

I'm obsessed with dancing. Specifically, Salsa, Bachata and Kizomba. Dance is obviously good for me in a bunch of ways; it's mild cardio, it's a creative outlet, is a large part of my social life, and of course it's a lot of fun. But it does result in late nights! Just how bad is it for my sleep? Let's find out.

ChatGPT agent has now come to people on the paid £20 a month plan. The shameless YouTube hype merchants are obviously impressed (must be hard having your mind BLOWN literally every few days...), others are disappointed, and others have found some decent use cases. I thought I'd put it to the test myself! Here I share my tests, and most importantly, the…

Last year I completed a six month research contract with The Good Growth Co. They're a brilliant social science think tank that serves the animal advocacy movement and alternative protein sectors in East Asia. We analysed a ton of social media data on how people across Southeast Asia are talking about meat, meat reduction and plant-based diets. To do this, we pioneered the use of…

I feel like every week I see the same debate play out again and again. "Is AI bad for the environment?" It a question on a lot of people's minds. I use AI for basically everything and so do many of my friends, so you can probably guess what my take is. But I've been a bit uncomfortable with how some people dismiss, or outright ridicule honest concerns about AI's environmental footprint.

When ChatGPT can understand human language baffling well, qualitative research could be set for a revolution. In the last few years there has been a flood of papers investigating the potential for Large Language Models (LLMs) like ChatGPT to revolutionise qualitative research. It can help with a variety of steps, from thematic code generation, to code comparison and refinement. It can achieve agreement with human coders in a fraction of the time.

A few months ago I gave a talk at the AI, Animals and Digital Minds conference in London. It was on the exciting new possibilities that have been enabled by AI deep research tools (I've also written about them here). The talk had an interactive workshop component where I had people generating deep research reports for themselves. We discussed tips, tricks and experiences.

Some people schedule meetings in order to avoid thinking. They have a problem at work and they either don't want to or can't figure out themselves. By scheduling a meeting with a bunch of other people, they can hash through it together. It gives them a kick up the arse to get something solved, that they put off on their own.

Recently, I gave a talk at VARC in Manchester on how we can make economic arguments against factory farming and for a plant-forward food system. I discussed a few reports I've written, some written by my team at Bryant, and some great reports by others. The talk was pretty well attended; I actually got last-minute bumped up to the main stage because so many people signed up, which I'm pretty happy about. The talk is accessible to most people; it's basically all pictures, with ZERO equations and almost no economic jargon!

Let's cut the crap: You have to make slideshow presentations all the time for work. You don't like it and would rather be doing actual work. I'm going to show you how you can use AI to make presentations in the fraction of the time. Advice tailored to privacy-concerned professionals Whilst there are brilliant new-generation AI presentation makers like Gamma…

I don't want to be dramatic, but feel like most organisations are thinking about AI wrong. They ask me: "Should I be using ChatGPT, Claude or DeepSeek?" "What are the best words to use in my prompts for fundraising?" What's the best AI tool for my job? They focus on tools. Sure, tools are important. But what is more important is…

"Richie, do you have any app recommendations?" It's a question I've been asked a few times in the last few months by people I coach. Following my policy of "any advice I give 5 times gets a blog post", here are 6 heroes of my workday. Superwhisper Speech-to-text has gotten a massive boost in the Generative AI era, and I love it!

It's barely April, and 2025 has already seen an AI breakthrough that I believe will change the work of nearly every office worker on the planet: AI Deep Research (DR) tools. If you've not heard of them, then stop reading this and go give them a try. You might just spit your drink out. If you've already spat your drink, and are desperate to get even more out of these tools: read on!

The International Energy Agency (IEA) thinks we should all chill out a bit.

Last reviewed and updated 23 march 2025 Since 2023 I’ve been giving free advising to non-profits on how they can leverage AI, particularly generative AI (ChatGPT, Gemini, Claude, Midjourney e…

Updated 16th March 2025 Perhaps about once a month someone messages me on LinkedIn or the Hive Slack community, asking how a data scientist can use their career to help animals. This is something I…

Last updated 10 March 2025 Making great graphs is hard. Any analyst can make a bar chart, but a good graph can hit home your point better than any text or table. A bad graph can take your reader in completely the wrong direction. Can we use AI models like ChatGPT to more easily make great graphs? Yes we can!

In 2012 we created 50 Things to do Before You’re 11¾ for the National Trust and the award winning campaign has since become their most successful behaviour change initiative to date. Designed to…

On 3 November, thousands of people joined the March for Clean Water to demand more action to clean up our waterways. They have a right to be angry: only 14 per cent of UK rivers are in good ecologi…

The CAWF report challenges the narrative that factory farming bolsters UK food security, revealing instead how it undermines it—costing taxpayers millions, threatening farmers’ livelihoods, and…

US-style mega-farms in Herefordshire face tough new regulations after high court ruling

Last updated 10th March 2025 Which AI large language model (LLM) should you use? With so many on the market, new developments every month and rampant cherrypicking when comparing models. It can be hard to know which is right for you. We have ChatGPT from OpenAI, easily the most popular, but we also have Google's Gemini model, Claude by Anthropic.