|
A.I. Zero To Hero In One Month - join me. Read time: 5 minutes Hey, welcome back. Last week, I talked about working til my teeth fell out, and rest. You can read that (and all past issues, here) This week, (high on painkillers still) I'm doing something I dont do. Its a sales email. And before I do - I want to make something clear - I dont sell ads in this newsletter. I am not against doing it - I just have no need to do anything that I dont think is of value to my readers. I've been offered money every week to run ads here and I say no. So what I'm about to do - is because I think it's valuable to you. I took my friend James' course - The Agent Growth Accelerator. I paid full price for it. I then sent 5 of my colleagues on it. I'm sending another 5 next cohort. I went from using the chat box pretty confidently to do tasks for me to building fully stacked agents that don't just speed up my work, they improve it. I built websites, dashboards, and more. In this newsletter I'm not going to convince you to do anything - I'll just show you some examples of what I did. And if you want to take James' course - go for it and you'll get a £100 discount when you use my code, that's the highest offer James has ever given (more than I got), because I told him I want my readers to upskill like I did. There's already a waiting list so I wouldn't wait longer, every cohort sells out and now I know why. I've never coded before Before you think I'm smarter/better whatever than you - think again. I switched from ChatGPT to Claude in January, figured out how to use Co-Work relatively well but never even opened Claude Code - I was scared of the words cos "I'm not a coder". In reality - you can get 10x your work done just working out how to use the new GPT work or Claude Co Work, and dont have to migrate to Codex or Code if you dont want to - but I do want to show you some cool stuff - to inspire you. I've never done a Cohort Course Before I'm also not a cohort or course junkie, but after several of my trusted impressive friends took James' course, I thought it might be time to give it a go - and the exceptionally clear agenda, well thought through guidance and most importantly - promise of where you'll get to in basically a month was compelling enough for me. In reality - I far exceeded my own expectations because I got lost in problem solving and my curiosity took me over the edge - but having trainers and support on hand during the course meant every time I was stuck, I could get unstuck. So, let me show you some cool stuff. Building A 'Second Brain' This is a fundamental starting point that unlocks everything - it's essentially moving your context of your life, work, goals, relationships etc into a folder on your laptop that exists locally, as a 'single source of truth'. Spending 2 hours setting this up properly pays dividends for the rest of your life - especially because once this is set up - you can move around to any AI model you want, just use the folder and whether its Gemini, Grok, ChatGPT or Claude - they proceed with all they need to do the job right - a huge unlock. The next level is to do this on something proper like 'Obsidian' using 'md' files cos they are tiny which means you use less tokens and speed up your whole AI journey - which is what I did. It's pretty fiddly, more complex, worth the effort but not for beginners - and where James' course is great is helping you on the right path - based on your own ambition here. Here's an image of my Second Brain Map to give you an idea of how it looks visually. What's great is, once this is set up, I created agents that automatically populate everything that happens into my second brain, building up more context - and then Claude can sort through everything methodically and surface what matters whenever. Security Obviously this was the most dry and boring part - but essential. Getting a fundamental understanding of what you are adding to your laptop, work environments and safely building is one of those 'learn it once, know it always' sessions no one would choose to do - but it came with a whole bunch of skills to download and put into your AI platform of choice, to make sure you never download or do anything you shouldn't. In many ways just this one session alone was worth the whole course price for peace of mind in the future, but it's boring, so let's move on to the fun bits. Adding Skills That Make Everything 100x Better I'm no gatekeeper - so whether you want to pay to go on the course or not - here are some of James' excellent skills on github you can download directly if you know what you're doing. Agent Foundation Skills To Download For Free Here You may have heard of skills too and not really understood what people are talking about or you may be pretty good with skills and ready to take it to the next level. Either way - this session came with so many brilliant skills, that all my work has improved exponentially - here are my favourite ones:
Once You've Got Skills, Time To Build Agents My favourites:
The bit that made it click: "Level 1 to Level 4" This is probably the part that blew my mind the most. The trick isn't building a clever agent. It's knowing how much rope to give it. James teaches this clever ladder with the agents
Some examples of things I built Once I'd got comfortable with the 'basics' - I started to explore moving from Co-Work to Code, to see if I could start building things, and wow, could I!? One of our biggest challenges at Heights is we work with lots of ambassadors, celebrities, influencers etc and the zillion people around them, agents, managers, etc... but a lot of the knowledge of pretty complex convos ends up in my head/inbox/whatsapp/insta DMs and its hard to surface - plus we have complex contracts that are always different, etc. However, to give the team the best chance of success - I wanted to build a really comprehensive site they could use so they can see everyone we work with, every detail on their contract, all leads, and even create a method for scoring new people we might want to work with. So I built the ambassador hub It's dynamic, sucks in all their meta data by the Windsor API so we can not only see how long is left on contracts but, how they are performing, spending, suggestions of what to improve, find out when they are up for renewal, proactive suggestions of what to say or do about content that's working (or not), and it does everything dynamically. So it helped me build a database for all the info - which can be edited directly in notion, or on the website directly, or if updates come in on slack/email - they will automatically populate that too. I taught myself how to host this on it's own website and for security reasons set up Google OAuth meaning only certain heights team members can access the info and only I can edit. From this base, I could add new stuff - my favourite part is this: Identifying Talent Using this system I was able to identify Alex Jones - the BBC One presenter - as a customer in our database and someone perfect for us to speak to a core customer, who wasn't currently being spoken to by us well. We get inundated with opportunities but its really hard to know who may or may not be a great fit for us. So I worked with Claude to design a process that enables me to stick in a celebrity name - and it will score them against our internal criteria, and check the ASA database to see if they are legit/challenging/brand safe, wikipedia, YouGov (which has a fame and popularity index for various celebrities - eg, Andew Tate is very famous but rightly very unpopular), and a few other places to help make an overall assessment - that comes out looking like this: Some Other Bits I also set up an 'agent email' inbox - so now I can send stuff to a dedicated inbox that's connected to my claude account and will get it to run tasks when I'm away from my laptop, and it emails my gmail with things like my morning briefing or my content suggestions. I instructed it to design it with HTML so it looks nice to read too - like this. I've also set up some direct agent messages to slack - but you dont need to get drowned in screenshots to make a point - the thing is I have exponentially improved my AI output from - OK to, actually pretty fucking stellar - in the span of a month. Oh and I helped my retail team visualise the dense data in a spreadsheet mapping out our sales in Boots retail across the UK with a visual dynamic dashboard. This took me about 15 minutes end to end, because once you start understanding how things work and connect - you start moving at pace problem solving. Health and other stuff I was going to show you the health stuff but actually this email is getting long - so maybe if there's interest we can go into that another time. Instead let me just end this to say - the only reason I bothered to hand craft and write this whole email (ironically no AI used other than to blur out private info) is because I want my loyal readers to grow with me - and if reading the science of success helps you discover something this valuable that I'm happy to write a whole email about - and you trust me as you do to guide you to a more successful life - then it was worth more than just one line. Yes James is my friend, but he's not asked me to do him a favour - lots of my friends have products or courses and I dont flog them anywhere - this entire email is one labour of love because I'd love you to feel how I do right now. Empowered as fuck. Ready to roll into this new AI world. Having gone from virtually zero to confident hero. With love and agency, |
Serial Entrepreneur and host of one of Europe's top business podcasts, Secret Leaders with over 50M downloads & angel investor in 85+ startups - here to share stories and studies breaking down the science of success - turning it from probability to predictability.
Refusing To Suffer Is What Makes You Sick. Read time: 6 minutes Hey, welcome back. Last week, I talked about how to know if your business idea Is worth pursuing. You can read that (and all past issues, here) Before we dive into the letter. Today I want to talk about a quote that’s been following me around for months and a pattern I keep seeing in myself and in almost every ambitious person I know. Carl Jung Said Something That Hurts To Read Read that again, slowly. "The foundation of all...
How To Know If Your Business Idea Is Worth Pursuing. Read time: 6 minutes Hey, welcome back. Last week, I talked about why you already know the answer to most of the things you're sitting on, and why nobody is coming to give you permission. You can read that (and all past issues, here) Before we dive into the letter. Hi from week 2 of my potentially terrible idea of a driving staycation throughout the UK. Staying over at Chester Zoo was a win, Chester is stunning, Liverpool was awesome and...
You Already Know. You're Just Waiting For Someone Else To Say It. Read time: 6 minutes Hey, welcome back. Last week, I talked about why the most important things in my life all came from chasing something else entirely, and the obliquity principle You can read that (and all past issues, here) Before we dive into the letter. I'm writing from week one of a two week staycation because with weather like this, why get on a plane with kids? Other than a few tantrums from the 5 year old and the...