Writeup

Nemotron 3.5 Super VlOpenCodehigh effortrun 16 Oct 202652,446 tokens

Instructions

Instructions — what the model was asked

Hey, I'm trying to write a post on my experience with Every Consulting. The consulting so far, I just joined as the head of tech consulting. I didn't tell you who runs consulting. I'm going to provide a transcript of me talking about my experience. I was also going to provide a tweet, which is just a list of the learnings that I've had so far for companies that were struggling to adopt AI. It...

Hey, I'm trying to write a post on my experience with Every Consulting. The consulting so far, I just joined as the head of tech consulting. I didn't tell you who runs consulting. I'm going to provide a transcript of me talking about my experience. I was also going to provide a tweet, which is just a list of the learnings that I've had so far for companies that were struggling to adopt AI. It was in response to Sam Par, who was asking, "How is everyone getting team adoption for Claude? How do you get your team to actually use it effectively without spending all their time on Twitter and learning?"

I need to write this in my writing style. I've included some writing of my own as a sample. I also have included a post on consulting at Every, so from in Natalia, you get a sense of what consulting looks like at Every. But yeah, focus mostly on the specifics of the interview that I provided with the context of that tweet.

https://x.com/hammer_mt/status/2032591631413567873?s=20

Learnings:

  • buy the model direct not 3rd party tools
  • raise the ceiling not the floor
  • workshops with 50%+ build time
  • assign impossible tasks
  • mandatory AI note taking + MCP connector
  • map workflows and systematically automate
  • train people to be managers (of agents)

In response to:
https://x.com/thesamparr/status/2032477081695584670
How is everyone getting team adoption for Claude?

I spent a lot of time on Twitter, as do you. We see all this AI stuff popping up. We're on top of it, or at least sorta. I know what's going on and are testing all these fringe ideas.

But how are all you people getting your team to actually use it effectively without spending all their time on Twitter and learning, which we know they won't and probably shouldn't be?

5 of 10 checks failed

ANSWER.mdthe model's output · 4.7 KBdownload

How to get your team to actually use AI

Sam Par asked a question I hear constantly: "How is everyone getting team adoption for Claude? How do you get your team to actually use it effectively without spending all their time on Twitter and learning?"

I joined Every a few months ago as Head of Tech Consulting. Natalia runs the consulting side. Our job is helping companies adopt AI without turning their employees into full-time AI researchers. The short answer: you don't mandate adoption. You make AI the only way to hit the goal.

Here are the learnings that actually work, from the work we've been doing at Every Consulting.

Buy the model direct, not third-party tools

People love evaluating wrappers — tools that use Claude or Codex or Gemini under the hood. But every wrapper bakes in someone else's product decisions and opinions. It's usually faster to build your own cloud skill with your own approach to work, and it's easier to configure and operate.

I don't know how a non-model company keeps up when Anthropic knows what releases are coming and schedules internal tools around them. Cursor is a great product, but how do you compete with Anthropic offering $5,000 of tokens a month for a $200 subscription? Third-party tools tend to be less flexible, less cutting-edge, and more expensive. Build, don't buy.

Raise the ceiling, not the floor

Most companies come in and say: "You all need to use AI now. We bought you the tools." That doesn't work. Even on pain of death, some people are emotionally unwilling to use AI. It's a self-preservation thing.

Use the carrot, not the stick. Nominate people who are already AI-forward and let them pull others along. Or borrow Boris Cheney's idea: slightly under-resource teams so the only way to hit the goal is with AI.

I make it explicit. If your goal is one blog post a week, you can do that manually. If it's one a day, you'll need AI in research, editing, everything. Don't set the goal as "starting today you produce one a day." Set it as "our goal is to get to one a day — what needs to happen?" When people know the target, they start thinking strategically about saving time with AI.

Workshops with 50%+ build time

If your AI workshop is all slides and no building, nobody learns. More than half the time should be spent actually building. People need the aha moment of making something work themselves.

Assign impossible tasks

Give people a task they can't complete without AI. Not as a threat — as a forcing function. When the task is impossible manually, they figure out the AI. That's where the real adoption happens.

Mandatory AI note-taking + MCP connector

Everyone on the consulting team uses Granola with the Granola MCP. It's the biggest unlock we have. After a meeting on topic X, I can generate an update email and send it to Y in seconds.

Roughly 80-90% of AI's value is extracting information from unstructured data and structuring it usefully. I pull meeting context into proposals and curriculum constantly. It's the kind of thing that makes AI feel like magic instead of homework.

Map workflows and systematically automate them

Take a repeatable process — like our discovery calls — map it out, and automate it piece by piece. Don't boil the ocean. Find the workflow, document it, then automate the parts that make sense.

Train people to be managers (of agents)

This is the one I keep coming back to. Individual contributors often resist AI because they enjoy the craft — they care how something is done. Managers have long since stopped caring how the problem is solved; they just want it solved to spec. Managers adopt AI more readily.

And here's the twist: managers are becoming individual contributors. A manager can manage a team of agents more effectively than a human team sometimes. It's lower bandwidth to vibe code a task in Claude than to brief a human, wait for them to use Claude, and wait days for a response. By the time you've written the prompt, you could have the answer in five minutes.

So we train people in management for the AI era: context switching, evals, developing taste for what to work on, and training AI in specific skills — like writing a really good PowerPoint skill or a daily update report skill for Claude.

The bottom line

You don't get team adoption by asking everyone to spend their days on Twitter learning the latest model. You get it by raising the ceiling, assigning work that requires AI, and teaching people to manage agents instead of just doing the work themselves.

If you want help figuring this out for your team, that's what we do at Every Consulting.


Note: The Natalia post on consulting at Every wasn't included in the files provided, so this post draws on your transcript and the tweet learnings, as requested.