25 Comments
User's avatar
Rich Carr's avatar

Daria! Readers: Do not skim past is one sentence in the middle: "Across all of this, my part is the judgment."

Two thousand users, three hundred workflows, agents running while you sleep, and the whole operation still routes through one human deciding what matters. That is the most honest description of working with AI I have read, and you wrote it in passing.

As for our time together, you did the hard part. You already knew everything about Amplifiers. You walked in carrying the whole album and walked out with the single. "AI's dream team, inside your AI" was in you the entire time, and watching it surface was the best hour of my week!

Daria Cupareanu's avatar

Rich, thank you so much for helping me find that clarity, everything makes so much more sense now!

David Jonah's avatar

I trust Daria implicitly and so I have registered for the webinar on August 05 so I can add another lever of advantage to my growing quill of arrows to tackle the most challenging part of being human which is, as Stephen Levy authored: "to speak in order to be understood and to listen in order to understand" Mind mapping indeed. Fascinating stuff

Daria Cupareanu's avatar

David, appreciate you 🙏🏼 you’ll have a blast

Rich Carr's avatar

You are going to absolutely love this based on these comments. Let me know if you're something you'd kike to unravel and we'll make it happen

TheAiBuildGuide's avatar

Rich, you're right to flag that line, it's the one that matters. But saying judgment matters out loud isn't the same as actually building a checkpoint for it, and that gap is where most of these break.

We've watched workflows run clean for months with a human "checking" only in theory, until the day nobody actually looked before something shipped. If that step isn't a real stop with something specific to check, it stops happening quietly and nobody notices until the wrong thing is already out the door.

Rich Carr's avatar

It's not lip service, ever. I've three questions that are maxims to ensure judgement, evaluation, and curiosity anchor it. 👍

TheAiBuildGuide's avatar

That’s the distinction I’d make too. The problem isn’t having a human in the loop, it’s having a human in the loop without a defined point of judgment. If those three questions are what force the evaluation to actually happen, that’s a very different thing from simply saying “a human is involved.” 🙂

Rich Carr's avatar

We're in lockstep

Alice E's avatar

This looks so helpful. I need to give it a try but I am not even there yet... I am in an uncertain place as regards what I am , let alone how to position anything on that basis! 🫣

Rich Carr's avatar

Alice, uncertain about what you are is the actual starting line, and it's where most people I work with begin. 🙂 Nobody hands you the answer. It's already in you, waiting for the right questions in the right order. Wednesday is free if you want to watch how that works. I hope to see you there.

Alice E's avatar

Okay

Think AI's avatar

A practical look at using AI as a team of specialists rather than one general assistant. The real value comes from repeatable workflows, the right tools, and human review before anything goes live.

Annabel Youens's avatar

Daria, I'm so intrigued by your Amplifiers! One part of me is - get them now! The other part is well, you haven't built them yourself so you don't know exactly what they do....I'm at a crossroads!

Daria Cupareanu's avatar

What are your main concerns? Quality of output or security and privacy?

Juan Salas-Romer's avatar

I build something useful then forget I own it three weeks later. The standing instruction to check once when the work changes is what I am borrowing.

Dallas Payne▫️'s avatar

Thanks for the shout-out, Daria! Such a fan of what you have created with Amplifiers!! Thanks for the lines to add to Claude's global instructions - that is super useful and I totally didn't think of that. This post is bookmarked to work through as I always find sooo many tips buried in your posts and this is no exception 😍

Daria Cupareanu's avatar

haha yeah, I'd say the biggest takeaway is to add that line, or a variation of it, to your global instructions - then you get to use Amplifiers more without having to remember Claude to use amplifiers 🤗

Mike Goitein's avatar

Can't wait to get started with the amplifiers, Daria!

Daria Cupareanu's avatar

& I can't wait to hear your thoughts on it!

George Shay's avatar

I like the marketing part of your product. How do you get access to it?

Daria Cupareanu's avatar

Hey George, you can connect Amplifiers to Claude or ChatGPT for free here:

https://getamplifiers.com/#setup

To access the PRO Amplifiers, you can either get a paid subscription to AI Blew My Mind, which includes PRO access automatically or start a free 7-day trial directly from the website.

John Brewton's avatar

The standing instruction to search the catalog before starting is the real system here, not the tools themselves.

jise's avatar

Thank you for the explanation.

Could you help me understand your role with Amplifier?

Is Amplifier a product that you built and actively maintain yourself, or are you primarily representing the product while a broader team builds and maintains it?

I'm also trying to understand the role of this newsletter. Is it your personal perspective as a builder, an official publication from Amplifier, or something in between? I found myself a bit unsure while reading.

Daria Cupareanu's avatar

My team and I built Amplifiers and of course, we actively maintain it. Since it’s my product, I talk about it here and share how I use it and the workflows I build for it. But my newsletter isn’t only about Amplifiers, you can see the rest of my articles here: https://aiblewmymind.substack.com/