We're moving from AI just knowing... to knowing AND doing

AI with ALLIE

The professional’s guide to quick AI bites for your personal life, work life, and beyond.

February 27, 2024

We're talking about AI buddies (copilots), smarty-pants bots (RAG), and agents that do the heavy lifting for us. It's about making our lives easier and our businesses smarter.

Also, an exclusive offer for my newsletter readers: use AIWITHALLIE to get $100 off my AI for Business Leaders course today.

So, if you’re ready to hear what’s next, read on 👇

Copilots, Bots, and Agents: Oh My

I went to a Microsoft hosted event in NYC last month that was attended by hundreds of developers. It ended with a fireside chat with Microsoft CEO, Satya Nadella and highlighted prompt engineering, building AI apps, AI code generation, data prep, gaming, and more.

The standout topics of the Microsoft developer event, the busiest sessions with lines out the door, were (1) building copilots and workflows and (2) RAG-based bots. These are technologies you can use today to redefine efficiency and accuracy in your business processes. A bit like mega advanced RPA but we’re going beyond automation and transforming how we approach problem-solving and innovation.

These two, plus AI agents, are stealing a lot of the tech buzz right now.

(And multimodal AI, but we’re not going to cover that in this issue because I’m typing so many social posts, my fingers are going to fall off my hand.)

Here’s what you need to know:

Can you explain RAG in crazy easy terms?

You bet.

Your data (like 50,000 internal wiki pages) is sent to a model (in this case, the embedding model). The model converts those 50,000 pages into numbers so an AI system can more easily navigate/search it. The number version of your data is then stored in a database. Cool. Then, you ask a question to whatever bot interface you’ve built. That question is also turned into numbers with the same number-turner model (again, the embedding model). The question-turned-number is used to search among the data-turned-numbers to find relevant sections or documents, like a data buddy system. Then a large language model (LLM) is given the full buddy system (the question + the relevant docs) to answer the question for the user. It’s like an open book test where the teacher is literally like “LOOK AT THIS PARAGRAPH ON PAGE 176 FOR THE ANSWER”.

Can you explain these three hot AI topics?

You bet.

1. Building Custom Copilots: Microsoft Copilots are AI assistants that are designed to work alongside humans on specific tasks. A copilot can sift through data, suggest solutions, and automate tasks. I built a Copilot during the Microsoft event and wouldn’t say it was the smoothest experience, so this is not me saying to jump on them. OpenAI’s version of these are called “GPTs”, which I do think perform extremely well on narrow tasks.

2. RAG Bots or Workflows: RAG (retrieval-augmented generation) combines LLMs with real-time external data retrieval and sourcing, improving the accuracy of AI responses and reducing hallucinations because of outdated info or limited context and allowing faster human validation. Again, sort of like an open book test in school.

3. Agents: Agents, or Agentive AI, is an AI that can take action. Adam Cheyer (co-founder of Siri who is featured in my AI for Business Leaders course) talks a lot about how AI is about knowing and doing. I forecast that in 2024, AI will get better at the knowing and going to start the doing by following documentation or instructions. We already see something like Hyperwrite’s AI Assistant in this space, which I’ll be demoing for you all soon.

Can you give me real-world examples?

You bet.

Here are examples for each ⬇️

Final thoughts

Individually, each of these offer an opportunity for enterprises to find methods for more AI integration into their ops. Custom Copilots offer a new level of productivity, RAG bots can bring higher reliability and accuracy, and AI agents can allow users to focus on more strategic activities. Together, they allow for a new level of autonomy and precision, enhancing how we currently use AI.

And as always: stay curious, stay informed,

Allie

My first cohort had hundreds of dedicated students from leading brands like Adobe, Mastercard, Amazon, Databricks, and more. We got an astounding average rating of 4.7 out of 5 stars, becoming the #1 AI for Business course. If you missed it the first time, now’s your chance to learn:

• How to become an AI leader in your org

• Finding, evaluating, and prioritizing real-world use cases and their trade-offs

• Strategies to enhance customer trust and value

• Path to responsible AI deployment and risk mitigation

On March 11, join me and four world-class guest speakers, including the COO of OpenAI, the Founder of Siri, an AI governance expert, and a 20-year tech and AI executive. Sign ups close March 10th. And now is your last chance to receive $100 off using code AIWITHALLIE100 until March 5th.

Tools, courses, and blogs that caught my eye:

Over the last few weeks, I’ve pulled together some of the top releases, and my take on each one. Check it out.

1. KitKat Ad — no, I’m not joking. KitKat Canada used a Google Deep Mind study to launch their AI-themed ‘Have a break’ ad (watch it + my thoughts)

2. Becoming an AI-Enhanced Human — my FREE 5-day educational email course on using AI tools to streamline daily tasks and boost efficiency overnight (with no experience required) (sign up)

3. Rabbit R1 AI Device🐇  — Rabbit launched their new R1 AI device at CES 2024 that seems like Siri on a beeper.. with a spinning toggle that nobody knows what it does? I bought it the second it came out, so your inbox will be the first to know when I receive it in Summer 2024 (read it) (my thoughts)

4. Andrej Karpathy’s Intro to Large Language Models — a tech-forward hour-long intro to LLM video from a brilliant computer scientist who just quit his role at OpenAI to do…who knows what (watch it) (my thoughts)

5. AI for Business Leaders course — if you liked my free email course, you’ll love my ‘AI for Business Leaders’ course. It’s the #1 AI Business course on Maven and has hundreds of happy students. Leaders from Disney, Apple, Google, Stripe, Oracle, Ralph Lauren, Chanel, CVS, Novartis, and JP Morgan Chase have already signed up. Get all of your AI leadership essentials in 40 video modules (sign up with discount)

6. Nocode.ai’s ‘A Guide to Retrieval Augmented Generation’ — If you want to read more about RAG integration, Armand Ruiz (my old coworker at IBM and a lovely human) discusses RAG essentials, its functionality, importance, operational mechanisms, and best practices for implementation (check it out)

Feedback is a Gift

I would love to know what you thought of this newsletter and any feedback you have for me. Do you have a favorite part? Wish I would change something? Felt confused? Please reply and share your thoughts or just take the poll below so I can continue to improve and deliver value for you all.