Hi, there!
Welcome to the 18th edition of Work in Beta.
In this edition, we answer the question we keep hearing: why does Claude feel different? Three reasons and an honest catch.
Also, if you are looking to build your individual or organisational system with AI, scroll down to the bottom of the newsletter to know more and connect with us.
We're planning a Claude Code workshop soon. Hands-on, practical, built for people who want to actually work differently with AI, not just watch someone else do it. Drop your details here to get on the waitlist.
Let's dive in!
IF YOU ONLY HAVE 2 MINUTES
THE ‘HOW TO’ PLAYBOOK
Claude Isn't Better. It Just Wastes Less of Your Time.
Something interesting is happening in our workshops and client conversations. People are fighting for Claude licenses.
Not because someone told them to. Not because procurement mandated it. They're going around IT, sharing accounts, burning through personal credits - because they tried it once on real work and now they don't want to go back.
In January, we asked our network which AI tool they use. 97% said ChatGPT. Just 14% said Claude. Two months later, we ran the same poll with our newsletter subscribers: 59% said Claude. 18% said ChatGPT. Same audience segment. Completely different answer.
But ask people why Claude and the answers get vague. "It just feels better." "The writing is more natural." "It gets me."
That vagueness is worth solving. Because the instinct is right, but the reasons are specific. Three layers specific.
What You’re Actually Noticing
Surface behavior gets you in the door. Verification burden is why people stay.
Plain language: how much time do you spend checking, rewriting, and second-guessing the output before you can actually use it?
We tracked this across 3 months of document-heavy work - proposals, contract reviews, competitive analyses. On average, we spend 40-50% less time editing and fact-checking Claude's output on these tasks compared to the same work in other tools. Not because Claude is always right. Because when it's uncertain, it says so - instead of writing something that sounds authoritative but is subtly wrong.
That's what we call confident nonsense - the failure mode that costs you the most time. Not AI saying "I don't know." AI writing something that reads like a fact but isn't one. You don't catch it on the first read. You catch it when a client emails you back.
The tool that creates less of that feels smarter, even when benchmarks say they're similar. And Anthropic has invested in this specifically - their long-context work showed a 30% reduction in incorrect answers and 3-4x fewer false claims on document-heavy tasks. That design priority shows up in daily use.
Why It Compounds
Projects remove cold starts. Skills turn corrections into reusable infrastructure - every fix becomes something you never fix again. Cowork operates on your actual files and folders, not just chat.
Together: the surface-level preference doesn't just persist. It compounds. Context and judgment accumulate instead of resetting every session.
We've written about this in detail: Skills, Cowork Plugins, Context Documents, Personal AI OS. If you haven't read those yet, start there.
The Honest Catch
Claude is not universally better. This is the part most takes leave out.
Claude out of the box, with no context, is just another chatbot. The advantage only shows when you do the setup work - Projects, Skills, context files. Without that infrastructure, you're paying more for roughly the same experience.
ChatGPT is still the most versatile generalist. Volume work, web search, rapid iteration, image generation, Custom GPTs, and no credit anxiety. For generalist tasks, ChatGPT does the job and does it well.
Gemini is increasingly strong if your organization runs on Google. Native integration with Docs, Sheets, Gmail, and Calendar makes it the natural choice for Google-native teams. And here's the thing most people miss: if your team collaborates inside Google Workspace, using Claude for individual work and ignoring Gemini for collaborative work is a mistake. The best tool for your work isn't always the best tool for shared work.
Claude's credit limits are real. You have to be strategic about when you use it. We see this as a feature, not a flaw - it forces you to reserve Claude for the work that matters. But it's a constraint you should know going in.
The infatuation is earned, but only for a specific slice: high-stakes, document-heavy, judgment-intensive tasks where tone and trust matter most.
The Mistakes We See People Make
Thinking Claude is universally better at everything. It's not. Other tools have real, distinct advantages. Claude wins on a specific kind of work, not all work.
Using Claude like any other chatbot. Blank chat, no Projects, no context. That's like buying a precision instrument and using it as a hammer. The advantage shows when you set up the workspace.
Burning credits on tasks other tools handle fine. Web search, high-volume brainstorming, rapid iteration - use the generalist for generalist work. Save Claude for the work that matters.
Expecting the feeling without building the workspace. The surface behavior gets you in the door. The verification advantage keeps you. But the compounding only happens when Projects, Skills, and context are set up. Without that, you'll drift back.
Final Thought
Claude isn't the make or break. No single tool is.
What actually changes your work is knowing which tool to use for which kind of task and setting up the one you choose so it doesn't start from zero every time. If that's Claude with Projects and Skills, great. If that's Gemini because your entire team lives in Google Workspace, that's the right call. If it's ChatGPT because you need volume and speed without credit anxiety, use it.
The people who are getting better output aren't loyal to a brand. They're loyal to a setup.
WORK WITH US
The Other 95%
Knowing how to prompt well is roughly 5% of what it means to actually work with AI. The other 95% - context architecture, workflow design, thinking behaviors, tool orchestration - is where your workday actually changes. Not "I got a better first draft." More like "I rebuilt how my team runs weekly reporting, and now it takes 12 minutes instead of 4 hours."
That's what we work on with professionals and teams through Work in Beta.
For individuals: We help you build the skills, context files, and workflows that turn AI from a chatbot into a system that actually runs your work.
For organizations: We help teams design the AI operating systems - skills, workflows, context architecture - that make adoption stick beyond "everyone has a login."
If you're curious what the other 95% looks like, reach out to us here.