AI's Not Magic, It's Method

i can’t tell you how many times an AI conversation with friends and family goes like this:

• “i’ve tried it… i wasn’t that impressed.”

• “it was fine… but it didn’t change my life.”

• “considering all the back and forth, my manual way is better.”

the problem with these statements?

they reflect bloated expectations of AI and its outputs.

and unrealistic timelines of how long it will take you to get them.

“if AI is gonna take all these jobs, shouldn’t it give me results like any other team member?”

first…

it’s okay to talk about AI without the additional discussion around it taking people’s jobs.

i’ll talk more about it in future editions. but for now, let’s just park it, shall we?

second…

you’re comparing your AI outputs to the ones you’ve seen online and in the news.

and you’re doing it without acknowledging that to get to that level you need inputs (prompts) that are:

• incredibly detailed

• created by experts

• and thoroughly tested

so today, let’s think of AI a little differently.

instead of an “all-powerful robot”, let’s think of AI as your newest company hire.

we’ll call him, Cal.

before we dig in, i want you to answer 2 questions below.

as you do, imagine that you’re Cal’s boss and he’s starting today.

1) what is your plan to set Cal up for a successful start?

• a) what plan? let's see what he's got.

• b) hand him today's tasks and hope for the best.

• c) take him through your formal onboarding process.

2) how much time have you budgeted to get Cal up to speed?

• a) pfft. Cal better come ready to roll on day one.

• b) a week at most, it's sink or swim around here.

• c) a month, minimum, before he's performing.

if you've answered anything other than "c" for both, it's time to start reevaluating your leadership 😐

it’s unrealistic to expect Cal to nail it on day one without guidance.

so, don’t expect AI to be any different.

instead of giving up after you input some bullshit and get a less-than-perfect result,

try these 3 simple tips to set yourself and your AI up for success:

1. give clear direction

2. share feedback

3. reset the chat

1) give clear direction: just like you'd guide Cal, get specific with your AI. define tasks clearly and provide step-by-step instructions.

use my prompt to see how:

[hey there - today i need your help creating a LinkedIn post using an outline + some rough notes. here are the steps we'll take together:

• i’ll share with you the post format i’d like to use

• i’ll share with you my outline + rough notes

• you’ll use this info to create a v1 draft

• i’ll review + provide clear feedback

• we’ll repeat until i’m satisfied

if this is clear, say "yes" and I'll proceed]

what this does:

• states my goal: “creating a LinkedIn post”

• outlines my inputs: “an outline + some rough notes”

• gives step-by-step instructions: the list of the steps we’ll take

• confirms my directions were clear: “if this is clear, say ‘yes’ and I'll proceed”

2) share feedback: AI, like Cal, learns from feedback. share it consistently to train and improve your AI over time. the more constructive you are, the better it gets.

use my prompt to see how:

[thank you for this. see below for a list of revisions to your v1 draft. i’ve included elements to keep as well as those to exclude in your v2 draft:

(include a detailed, point form, list of revisions, inclusions, and exclusions for the v2 draft)

please respond by sharing a v2 draft with this feedback reflected.]

what this does:

• states my goal: revisions to the v1 draft

• outlines my inputs: detailed, point form list of revisions, inclusions, and exclusions

• clarifies the next output: a v2 draft of the post with this feedback reflected

3) reset the chat: i use many of my AI chat windows for specific, recurring, tasks (ContentEngine, NewsletterEngine, ProposalEngine et al.). this allows me to train my AI on the specific tasks, based on my process and desired outputs for that work.

however, just as Cal would be overwhelmed by too much information, so too might your AI.

so, if it's been a while since you last used a specific window, give your AI a quick catch-up and reminder of your previous work.

use my prompt to see how:

[before we begin today's tasks, please review our chat history to familiarize yourself with our work to date. this context is critical in your performance of today’s tasks.

once you have finished, say “i’m ready for my next task”.]

not bad right?

1 mindset shift and 3 simple prompts to help you get the best results and outputs from your AI.

so, as you head into next week take this one tip with you…

clear instructions, regular feedback, and consistent resets are 3 simple things you can do TODAY to help you unlock AI’s true potential in your life and business.

AI is a machine.

built on complex models and BILLIONS of pieces of data.

you need to onboard it. you need to train it.

just like it’s your newest hire, Cal.

that’s all for today beautiful humans.

keep pushing and remember…you got this 💪🏻

see you next week,

kristina, with a “k”

BEFORE YOU GO! PLEASE TAKE 30 SECONDS FOR THE ONE-CLICK POLL BELOW TO SHARE YOUR THOUGHTS ON THIS WEEK’S EDITION.

P.S.

• i’m Canadian. there will be extra “u”s, double “l”s, and more. this is not a mistake. it’s willful.

• i’m Queer. a mother, a wife, and an advocate for equity of all people, not just Queer folks. i openly, and without filter, share my own personal narrative and life’s stories.

if any of the above isn’t for you, please unsubscribe. no hard feelings.