www.imarketing.

courses
www.imarketing.courses

8020 AI Hyperdrive Session #2

Brought to you by:

Samuel J. Woods

© Samuel J. Woods. All rights reserved. 1

 www.imarketing.courses
www.imarketing.courses

Terms of Use
Copyright © Samuel J. Woods (the “Author”) in perpetuity. All rights reserved.

You may not republish, reproduce, duplicate, copy, sell, create derivatives, display, disclose,
distribute to strangers, colleagues, friends, family, or any other third party, or otherwise use any
material from the Program or Content for commercial purposes or in any way that earns you or
any third party money (other than by applying them generally in your own business). By
downloading, printing, or otherwise using the Program or Content for personal use you in no
way assume any ownership rights of the Content – it is still the property of the Author. Any
unauthorized use of any materials found in the Program or Content shall constitute infringement.

That means, for example, that you may not copy or create derivative works of any material, and
then sell them as your own invention.

Pirating or reselling is expressly prohibited and will be prosecuted by law, including seeking
damages and recouping lost revenue.

No part of this document may be reproduced or transmitted in in any form or by any means,
electronic or mechanical, including digital copying, photocopying, recording, or by any form of
information storage and retrieval system, except by a reviewer who may quote brief passages in
a review to be published on a website, email or social media, without permission in writing from
the Author. If you reference or quote, please give credit to: “Samuel Woods”.

Everything in this document is provided as-is and does not constitute general or specific advice
on what you should or should not do in your business. You’re responsible for applying any
material as you see fit and cannot hold the author responsible for any way you’re using the
material.

Note from Sam: Over the past few years, I’ve seen my work copied, pirated, stolen, and
presented as someone’s own with zero credit. I can’t stop or prevent it and the overwhelming
vast majority of my students and clients, like yourself, don’t do that. So, I’m asking you to please
don’t be one of the very few who steal or pirate my work. I’ve spent years developing these
heuristics, models, strategies, tactics, in many cases actual tools, and numerous prompts. I’d
rather see you use them inside your business than anywhere else.

© Samuel J. Woods. All rights reserved. 2

 www.imarketing.courses
www.imarketing.courses

#1. Skeleton Key for AI Tools

The most important 80/20 principle for tools is:

What you gain in convenience, you lose in control.

The more you rely on 3rd party SaaS, apps, and tools to have work done for you, the
less control you have over input (like prompts) and output.

© Samuel J. Woods. All rights reserved. 3

 www.imarketing.courses
www.imarketing.courses

This is not a bad thing. It just means you, as the Chief Robot Officer in your business,
must take into account how much control you need to get the job done.

This also means you have to decide how much of your data (by data, I mean everything
from knowledge, actual data, information, etc. that you have in and about your
business) you want possibly exposed to a 3rd party.

A general rule of thumb is:

If the output will become public at some point (like ads, emails, content, etc.) then, for
example, using ChatGPT and GPTs from OpenAI could be the right choice for your
business.

But, if you have proprietary data of any kind that is unique to your business, then you
should seriously consider not having it exposed to a tool like ChatGPT.

For a lot of tasks in your business, you can have them done with ChatGPT (and GPTs)
or Poe.com (the paid version has Bots that function much like GPTs, by the way).

Claude.ai is another model that’s great for writing, for example.

The most important thing to remember is:

Your output is as good as the CONTEXT you provide.

Context, if you recall, are things like examples of what you’re looking for, knowledge
files, criteria, information about your business, products/services, your avatar/persona,
etc.

You can have the simplest prompt in the world (“Write 10 new LinkedIn posts for me,
based on the examples I’m giving you”) produce excellent output—only because you’ve
provided good examples.

With that in mind, the next principle to keep in mind is:

All tools are sequences of prompts and tasks.

By all tools, I mean anything from Saas or Apps available, to Bots, AI Agents,
AutoGPTs, Zapier, N8N, Make.com and so on.

© Samuel J. Woods. All rights reserved. 4

 www.imarketing.courses
www.imarketing.courses

(For clarity, people are using different terms for the same thing: “AI Agents”,
“AutoGPTs”, “Bots”, etc. They’re the same: a sequence or prompts and tasks, often
done via API calls or automated steps).

You can have the same task or job performed in any which way:

Prompts: Tell ChatGPT or other LLMs what you’re looking for and want it to do.

Tools: like scrapers, Zapier, N8N, Advanced Data Analysis, etc. that perform
parts or most tasks.

Bots: like Agents or GPTs that automate parts of the workflow.

Apps/Software: A series of API calls (prompts and tasks) to different tools, like
GPT-4 and GPT-4 Vision that do it all for you.

Again:

No matter what specific tool you use, they’re all a sequence of prompts and tasks.

Therefore, the 80/20 of AI tools is for you to have clarity on your job process, so you can
turn it into a sequence of prompts and tasks.

Once you have this, mapped out, you can turn it into anything you want (prompts you
use manually, a Zapier or Make.com flow, or a Bot, or something else).

As the Chief Robot Officer, you need to evaluate what’s best for your business, in
terms of costs and effort involved.

For some businesses, using a 3rd party tool like Jasper for marketing content is the best
choice.

For others, you may prefer your own Zapier sequence. Or, all you need is a GPT inside
ChatGPT to get the job done.

To make the decision, you evaluate:

● Your data: How much or what data can be “exposed” to a platform like OpenAI
vs. what you need to keep private.

© Samuel J. Woods. All rights reserved. 5

 www.imarketing.courses
www.imarketing.courses

● Your job and use cases: What is the most cost effective tool to get the job done?
Do you really need to spend the time and effort to set up a Zapier sequence? Or
can you simply make a GPT?

If you can pay a human $250 or less to do a job, that job can be done by AI (either
completely or in parts, for now).

What I’ve given you here is the fundamental understanding of how this works, so you
can decide what’s best for your particular business.

Again:

What you gain in convenience with tools (prompts, GPTs, Zapier, etc.), you lose in
control (of the output or outcome quality).

The closer you are to an LLM, the more control you have.

LLMs and Machine Learning models (what everyone is calling “AI”) are powerful.
Because it’s not about simple automation of jobs and tasks.

Instead, think of AI as “Automated Intelligence”.

Jobs and tasks that require the “human element” can be augmented, improved or even
replaced now with LLMs (or other ML, vision, images, or audio models).

We’ll cover more on this in the next segments of this session and over the upcoming
Sessions.

For now, as a starting point, I’m providing a simple stack of tools you can get started
with. These are options, not endorsements.

© Samuel J. Woods. All rights reserved. 6

 www.imarketing.courses
www.imarketing.courses

AI/ML Business Stack

Directories
There are countless directories for AI tools and more popping up every week. Here are
3 you can start with (and you likely don’t need many more).

● Supertools: https://supertools.therundown.ai/

● FuturePedia: https://www.futurepedia.io/

● FutureTools: https://www.futuretools.io/

General purpose for < $250 Tasks

● ChatGPT Plus (includes GPTs)

● Poe.com (paid includes GPTs, multiple models)

● Claude.ai

● Bing (Copilot)

● Perplexity.ai

● Julius: https://julius.ai/

GPTs for ChatGPT

● All GPTs: https://allgpts.co/

● GPT Dex: https://gptsdex.com/

Agents, Bots, Automations

● Zapier: https://zapier.com/

● N8N: https://n8n.io/

● Make.com: https://www.make.com/

Assistants

© Samuel J. Woods. All rights reserved. 7

 www.imarketing.courses
www.imarketing.courses

● Hyperwrite: https://www.hyperwriteai.com/personal-assistant

● Embra: https://embra.app/

● Harpa: https://harpa.ai/

● AiAgent: https://aiagent.app/

“Talk” to your data, aside from GPTs or Bots via Poe.com:

● Hyper: https://gethyper.ai/

● Hansei: https://hansei.app/

Marketing and Sales

● Jasper: https://www.jasper.ai/

● Flowpoint: https://flowpoint.ai/

● Crystal: https://www.crystalknows.com/

● Mixo: https://www.mixo.io/

● Postwise: https://postwise.ai/

Video

● HeyGen.com

● RunwayML.com

● Fliki: https://fliki.ai/

● Klap: https://klap.app/

● Pika: https://pika.art/

© Samuel J. Woods. All rights reserved. 8

 www.imarketing.courses
www.imarketing.courses

#2. Rosetta Stone for Prompts

The most important 80/20 principle for prompts is:

Prompts are simple instructions in natural language, the way you would talk to a
human.

From this principle, you can extrapolate and say:

© Samuel J. Woods. All rights reserved. 9

 www.imarketing.courses
www.imarketing.courses

Good prompting is simply good communication. Bad prompting is simply bad

communication.

If a human can’t understand your set of instructions, then an LLM will struggle with it
too.

In other words:

Talk to an LLM like you’d talk to a human being.

This is partially why ChatGPt took off like it did. It made the interaction with an LLM a
chat session, the way you’d chat with or text a friend.

This is another 80/20 principle of prompting:

You will always get better output if you engage in a conversation with an LLM,
versus going in “cold” and firing off a single prompt.

Which brings me back to what I said in Session 1:

What would someone need to know, and what information do they need, to do
their job well?

Whatever that is, that’s what you bring with you to your prompt (or a sequence of
prompts).

Quite honestly, if you have a document with a set of instructions that you’d give a team
member, for any given task—this is your set of prompts that you use.

You turn task descriptions into a sequence of prompts.

Which bring us to another 80/20 principle of prompting:

Language is nearly infinite. You can express the same message in multiple ways.
Therefore, prompts are also endless.

What matters is how clear your intent is, not necessarily the exact words.

Other reasons why there’s not a perfect combination of exact words that will always give
you the best output:

© Samuel J. Woods. All rights reserved. 10

 www.imarketing.courses
www.imarketing.courses

You can use the same exact prompt ten times in a row, and get variations, nuances, and
differences in the output ten times. You’ll never get the same exact output.

Why?

Because LLMs are narcissistic.

They “prefer” their own words over yours. And never, ever follow the exact letters and
words in your prompt.

But LLMs follow the direction you give, in the intent.

They turn your words into numbers and never preserve the exactness of your words.

Therefore, there’s not a single perfect prompt anywhere. It’s not even technically
possible because of how LLMs treat input and output.

What matters is the intent and direction of your words.

It’s the same as with humans and language:

If a human does not understand you the first time, you change what you said but
preserve the intent.

Some formulations of a message are better than others. But remember what I said
earlier:

Talk to an LLM like you’d talk to a human being.

And as with good communication, there are properties to your message that will make
your prompts the best they can be.

Here’s a universal LLM prompt template you can use with any LLM.

© Samuel J. Woods. All rights reserved. 11

 www.imarketing.courses
www.imarketing.courses

Universal LLM Prompt Template

FUNCTION / Role: Describe the role and function of what you’re looking for.

Task: Describe the specific task or job you want the LLM to do for you.

CONTEXT: Provide details, information, data, knowledge, instruction, examples that

are relevant to the task.

MODIFIERS / Rules: Provide specific rules you want the LLM to follow in how it
treats your input and what kind of output you want.

© Samuel J. Woods. All rights reserved. 12

 www.imarketing.courses
www.imarketing.courses

#3. Being the Chief Robot Officer

As The Chief Robot Officer, you direct a

strategic division of labor switching between
AI and human tasks (Collaboration),
assigning responsibilities based on the
strengths and capabilities of each (Bionic).

You work with:

● Heuristics
● Mental Models
● Strategies

And you oversee the Regenerative

feedback flow, where you distribute pieces
where they could help make other work
faster (Optimize) and better (Maximize).

This is truly the end of $10/hour work. This no longer exists.

I’m even going to say that any job that’s $250/H or less is either fully replaced,
automated, eliminated, or can be done in part by AI—right now.

Why is this true?

Because if any part of your work touches a computer or the internet, or both, then it can
not only be automated as tasks but intelligence is now outsourced

We’re going from automating clicks and tasks (do this, click here, send this message,
perform this function) to now integrating and automating intelligence, knowledge, and
cognition.

All $10-$250 tasks can now be automated, augmented, eliminated or delegated.

© Samuel J. Woods. All rights reserved. 13

 www.imarketing.courses
www.imarketing.courses

As the Chief Robot Officer, your role is to look at jobs and tasks and ask:

● Augmentation: Can this task be enhanced and done better with a human +
AI/ML tool?

● Automation: Is this task routine and repetitive?

● Delegation: Can this task be delegated to fully automated, automation plus

human, or does it still require a human for most or all of it?

● Elimination: Is this task even necessary? Can we eliminate it? Automate it?

And this is far reaching, into any job.

If the job is clicking buttons, using software of any kind, or working with any kind of
information (from text, to data, to images, to audio, video, anything), AI can do it in parts
or completely.

LLMs have the capability to draft various types of documents, including code files, legal
documents, and scientific documents, based on given instructions.

LLMs can already surpass human capabilities in various aspects of learning, “thinking”,
and doing.

They don’t learn, think, and do in the way humans do.

But they emulate, mimic, or do it in their own way (which AI researchers aren’t crystal
clear on exactly how yet).

LLMs already at the full stack of mental capabilities that humans have, as they can
generate new and original work.

LLMs like GPT-4 can connect dots and synthesize new information, even if it wasn't
explicitly in the training data.

The ability of language models to use entirely novel information and improvise is a sign
of their own kind of intelligence and creativity.

With that in mind, how do you decide what is done and by who?

© Samuel J. Woods. All rights reserved. 14

 www.imarketing.courses
www.imarketing.courses

If your job or task involves ANY one or many of the following, it can be Automated,
Augmented, Eliminated or Delegated:

● Analyze
● Plan
● Organize
● Coordinate
● Communicate
● Solve
● Evaluate
● Monitor
● Improve
● Innovate
● Research
● Write
● Information

You take stock of what your job or task consist of, and if it involves any of the above, an
AI can do it fully or in part.

With that in mind, this is why you need to guide your integration by the Heuristics I’ve
shared so far:

● Bionic (Human and Machine).

● Collaboration (Not just separately but together).

● Regenerative (One thing feeds into another).

In other words, you ask:

● Bionic: What of all the parts of my job can be done fully by a Machine? What still
requires a Human?

● Collaboration: What parts must be done in collaboration?

● Regenerative: What parts can contribute to other parts, either in the same task
or in other parts of my business?

And then, once you have a list of those parts, you ask yourself:

© Samuel J. Woods. All rights reserved. 15

 www.imarketing.courses
www.imarketing.courses

If this AI was a human, what would s/he need to know, in order to do their job
well?

This becomes the Context and Prompts you provide.

Let’s look at the 80/20 Model I shared in Session 1:

Analyzing an email campaign can be Bionic (some things done by the Machine, other
things done by a Human).

Take, for example, data analysis:

I would upload campaign data and metrics to something like ChatGPT and then ask it to
perform an analysis of the performance (I showed an example of this in Session 1, go
back and rewatch if you need to).

The human side would be to review the analysis and keep asking questions for further
analysis.

The kind and types of prompts I would use, would be to analyze and explore the data:

For the next step, it’s all about Collaboration:

© Samuel J. Woods. All rights reserved. 16

 www.imarketing.courses
www.imarketing.courses

This is continuing the conversation with ChatGPT about my email performance, where I
probe deeper in Collaboration with ChatGPT, and the kinds and types of prompts are:

Are you starting to see how this Model from Session 1 is making sense, as a
foundational 80/20 principle?

It’s a continuing loop:

© Samuel J. Woods. All rights reserved. 17

 www.imarketing.courses
www.imarketing.courses

#4. Additional Perspectives

Evolution of Understanding
Our understanding of anything evolves from Hypothesis and goes all the way to
Wisdom. Most people never go that far, or even achieve Insight. In the Age of AI, you
must evolve to Insight and Wisdom.

© Samuel J. Woods. All rights reserved. 18

 www.imarketing.courses
www.imarketing.courses

From Awareness to Wisdom

AI models can be Aware, exercise Cognition, Intelligence, and Knowledge. To some
degree, they can also provide Insight.

Represents the pinnacle of cognitive and emotional

development. Incorporates elements of knowledge, experience,
Wisdom
and deep understanding, with the ability to make sound
judgments and decisions in complex situations.

Deep, intuitive understanding of a person, thing, or situation.

Insight Leads to innovative ideas and solutions and is critical in
understanding complex problems.

Collection of information and skills acquired through experience

Knowledge or education. More factual and less abstract, essential for
developing understanding and insight.

Ability to learn, reason, and solve problems. Fundamental

Intelligence cognitive capability underpinning the acquisition of knowledge
and development of insight.

Mental processes involved in gaining knowledge and

Cognition comprehension, including thinking, knowing, remembering,
judging, and problem-solving.

Most basic level, involving consciousness of one’s environment,

Awareness oneself, and one’s own thoughts and feelings. Starting point for
all higher cognitive processes.

© Samuel J. Woods. All rights reserved. 19