- Everyday AI
- Maybe the most important newsletter from us 👇
Maybe the most important newsletter from us 👇
✅ Hundreds of hours of GenAI insights in 5 steps for business implementation, Neuralink's first brain chip implant, overview of ChatGPT @mentions feature, and more!
🦾How You Can Leverage:
The company that figures out AI implementation is gonna win 2024.
Lucky for you: we have the blueprint here.
The newsletter is definitely the tl;dr version of today’s show, which should be a requirement for anyone interested in GenAI.
Like we mentioned above, a lot of the content from today’s show goes out to some stellar guests who shared their insights on (and off) camera.
We’ve had 120+ amazingly talented guest on our show, and have worked offline with thousands of others teaching the basics of GenAI.
While the majority of today’s show came from our own experiences helping clients implement GenAI, we spent hours digging into insights from these amazing people and previous episodes 👇
Ep 146: IBM Leader Talks Infusing GenAI in Enterprise Workflows for Big Wins with Ben Mandelstein rom IBM
Ep 116: Using Gen AI in the Enterprise Space – Insights From Walmart with Jack Adams from Walmart
Ep 179: Mastering Prompts With An OpenAI Ambassador – The One Secret Skill Revealed with Abran Maldonado from OpenAI
Ep 159: Using Microsoft Copilot To Advance Your Learning and Leadership with Carol S. Scott from Microsoft
Ep 97: Combining AI + HR: How to do it responsibly with Jen Kirkwood from IBM
Ep 23: Using AI in the C-Suite for Human Connectivity with George Alifragis from the Cyber Security Global Alliance
Ep 79: The Importance of Ethics in AI with Moe Alo from IBM
Ep 90: How To Tackle AI Privacy and Governance with Katharina Koerner from the Tech Diplomacy Network
Ep 145: NVIDIA Leader Talks GenAI + Data: Unlocking new ways to interact with our world with Adam Scraba from NVIDIA
Ep 148: Safer AI – Why we all need ethical AI tools we can trust with Mark Surman from Mozilla
Ep 98: The Art of Storytelling – Experience From a Microsoft AI Ambassador with Doug Thompson, former Microsoft AI Ambassador and Owner of The Doug Thompson LLC
(See that ChatGPT? You gotta cite your sources shorty!)
Whew. That’s a lot of additional listening. If you had no plans this weekend, now ya do!
Now, let’s dig beneath the sand and find those golden AI nuggets, shall we?
This is the legit blueprint on how to implement GenAI in 2024.
1 – Gather insights from a ground-up committee️
You don’t start with guidelines and guardrails.
You start with a fast-paced, yet intentionally deep AI committee built from the ground up.
Tackling tough questions before anything is set in stone is important for widespread buy-in and transparency through a process that is honestly a bit uneasy for everyone.
Start by asking Why GenAI.
(The answer might be based on studies that show GenAI can automate up to 70% of what employees spend their time on today)
And talk openly about what happens if AI works. (More on that in #5)
Try this: Your AI implementation committee should have more entry level workers than upper management, directors and C-suite. If you want to be successful, you have to thoughtfully take into account insights from the very people who may be using GenAI the most on a daily basis.
(Hint, that’s probably not the C suite. If you build top-down with mandates from the mountain, you’ll fail.)
2 – Create straightforward guidelines with guardrails
Imagine being in a crazy car.
But it’s a completely foreign type of vehicle on some otherworldly road.
All those decades of driving won’t help you there, right?
That’s what Generative AI is like once your company fully embraces it.
Hence, the need for guidelines and guardrails.
Before you can sprint toward your first implementation, you need to take all of the insights from your (deliberate yet quickly moving) AI Committee into rules on paper.
Try this: Again, don’t fret shorties. So many great orgs have already put great rules in place that you can borrow from. (Don’t steal em, but use them as your inspo.)
Oh, also check to see what existing policies you already have in place that could be updated to include provisions on responsible and ethical AI use.
3 – Sprint toward your first small measurable AI project
If you think your companies first AI project should be a one-year pilot program building your own Large Language Model, that’s a recipe for failure.
Instead, we suggest focusing on an easy win that makes sense across all departments.
We talked about 3 examples, and gave recommendations on how to communicate the story to other departments and stakeholders.
Try this: Make sure your first AI implementation makes sense and is quantifiable. Can’t emphasize enough how important that piece is if you wanna win in the long-run.
4 – Invest heavily in education and training that align with long-term business goals
This one’s a doozy.
To be honest, proper GenAI implementation might mean unlearning decades of best practices.
The best piece of advice here is focus on ongoing education and GenAI learning opportunities for your company. (We do that for individuals and companies! Just reply to this email for more info.)
Before you can properly leverage GenAI in the long-term, you have to focus on explainability.
You can’t just let the black box of Generative AI remain a black box. To increase trust, you need explainability.
Try this: We took a DEEEEP dive into this one yesterday, so make sure to check that show out.
5 – Plan for a future of what happens when AI works
What happens when AI actually works?
The answer is hard to come by, actually.
You know that whole ‘83% of companies say GenAI is a top priority, but only 4% have implemented GenAI across their entire org.’
But if you do GenAI implementation the right way, you have to prepare for what comes after. When you follow steps 1-4, how your company does business changes completely.
Try this: Before you’re fully off to the GenAI races, have the tough conversations. Do jobs go away? Do impacted workers get reskilled and upskilled? Will job roles change?
Those questions are for your company to figure out, but they’re required questions to explore.