AI, Jobs, and the Surprising Lesson We Can Learn from Horses 🤖🐴

Picture this: two horses chatting in the early 1900s. One's all jittery about these newfangled "mechanical muscles" (read: cars) making them obsolete. The other's like, "Chill, dude. Remember when we had to run coast-to-coast delivering mail? Or that time we were in that big battle? City jobs are a breeze! Even if cars become a thing, we'll find other gigs." But, spoiler alert: it didn't pan out that way. By 1915, the horse job market was in a freefall.

Now, here's where it gets interesting. "There isn’t a rule of economics that says better technology makes more better jobs for horses." Sounds bonkers when you put it that way, right? But swap out "horses" for "humans", and suddenly it's a sentiment many nod along to. Just as machines replaced horses, robots with their fancy decision-making skills are gearing up to give us a run for our money. It's not about them being stronger or faster; it's about them thinking and deciding. And trust me, they're doing it at a pace that makes our biological evolution look like it's stuck in traffic.

You might be feeling all smug, thinking your job is irreplaceable. But remember, technology doesn't sleep. It's like that one friend who's always up for a midnight snack run – it just keeps going. The car was the beginning of the end for our equine pals, and now, ironically, it's hinting at our future too.

So, what's the takeaway? Don't be that second horse, thinking things will always stay cushy. The future's coming at us fast, and if we don't saddle up and prepare, we might just get left in the dust. And hey, while we're at it, maybe it's time to be a bit kinder to our robot pals. After all, they might be our coworkers sooner than we think! 😉🤖

From Single-Use Gadgets to Universal Craftsmen: How ChatGPT is Revolutionizing Robotics and Beyond

I came across a great AI article the other day that revealed a major insight about the future of ChatGPT and generative AI.

The article quoted an AI startup founder who said they thought they'd have to seriously customize ChatGPT to work for robotics. But surprisingly, in many cases, they found that plain old ChatGPT could control their robots pretty well straight out of the box!

As someone who geeks out on all things AI, this blew my mind. 🤯 It suggests AI models like ChatGPT are becoming powerful general-purpose tools, not just niche applications.

The article went on to explain how we got to this point. So let me break down the key points for you all...

Until recently AI models were like those single-use kitchen gadgets that only slice mushrooms or squeeze limes. They could only do one thing. For example, Google's AlphaFold AI just predicts protein structures.

But the founders realized they could just use ChatGPT to control their robots without training it specifically for robotics at all! I've heard similar stories from people in healthcare, semiconductors, insurance - you name it.

So how did we get to this amazing point where AI can do almost anything?? Let’s break it down….

First, a quick AI 101. These large language models like GPT3 and GPT4 work using probability. They look at an input and predict the most likely next words. So they're like a supercharged autocomplete. But instead of just right and wrong, it's about more likely and less likely.

This is powerful but unpredictable - it can make weird mistakes. But it also means AI can be flexible and creative in awesome new ways! We just need to corral the randomness.

Here's an analogy: Before quantum physics, scientists thought things behaved predictably. But the quantum world is random! At first that blew their minds. 🤯 But then they learned to harness quantum randomness in useful ways, like building semiconductors and computers!

Same with AI - instead of fighting the randomness, let's load the dice in our favor.

How do we load the dice? With something called gradient descent. This lets us take a bad AI model spitting out gibberish and slowly train it to get better and better. Like teaching a kid to ride a bike - they wobble at first but get better with practice. Gradient descent is a game changer because we can take an existing AI and fine-tune it by showing it new examples, like a guitar player learning new songs.

So with that foundation, let's look at two key innovations that led to ChatGPT:

1️⃣ Teaching AI to follow commands:

Early AI models like GPT2 were like talking to a toddler - they would randomly babble based on the prompt you gave them. Not so useful!

But GPT3 was trained to actually follow instructions and answer questions. Now you could say "write a short blog post about cats" and it would deliver. Way more useful!

It's like teaching a kid to follow instructions vs babbling randomly. This training breakthrough let GPT3 take direction instead of doing its own thing.

2️⃣ Teaching AI to take feedback:

Following commands was great but still limited. It was like a slot machine - pull the lever, and hope the output was good. But as any writer knows, creation is an iterative process!

So for GPT3.5, they trained it to accept feedback and iterate, like a back-and-forth conversation. Now you could say "here's what you wrote, please add more details about the cat's fur." And it would build on its previous output.

This was huge - now humans could guide the AI from bad to good to great! More slot machine, now it's like golfing - each swing builds on the last to get closer to the hole.

So with these breakthroughs, ChatGPT was born! They combined command following + feedback with conversational interfaces. But here's the big insight: these models can now be used in totally new ways beyond just chatting.

Think of the AI model like a computer processor - it takes input data and processes it. But instead of code, we can give it natural instructions and data, and it can handle text, images, code, you name it!

Let me share some real-world examples:

Health insurance - Instead of training an AI with all your policy docs, give it specific docs + a question and have it answer based on those. More secure and customizable!

Robotics - Turn actions into a "code" language. Have AI remix existing programs into new ones for new tasks. Like a Zork adventure game!

Chips - Use AI to remix and modify chip designs faster. It's like autocomplete for hardware!

Wild right? We turned AI from a specialized tool into a versatile crafting material, like turning wood into paper. The key is focusing less on AI as an independent agent, and more as a programmable tool.

So in the future, instead of specialized AI for everything, we can use these models as "anything tools" for all kinds of tasks. The democratization of AI!

What do you think? Let's chat - are you excited about the new possibilities with AI? Or worried it's going to take over the world?

Unlocking Time with AI: 6 Game-Changing Ways Businesses & Employees Are Reclaiming Hours!

Hey there, fellow time-travelers! No, I'm not talking about hopping into a DeLorean and zipping back to the '80s. I'm talking about the modern-day magic of AI, a tool that's letting businesses, especially SMEs, reclaim the one thing we can't get enough of: time. After all, time is more valuable than money.

AI: Your Personal Time Machine

1. Bored of Boring Tasks? Automate 'Em!

Remember those long nights spent on data entry or scheduling? AI's got your back, taking over these yawn-inducing tasks. Now you can channel your inner Da Vinci and focus on the creative stuff. Or, you know, catch up on your favorite Netflix series.

2. Decision Making: Fast and Furious

With AI, you don't need to be a Sherlock Holmes to make quick and accurate decisions. It's like having a supercomputer in your pocket that does all the heavy lifting. No more endless hours of data analysis. Just pure, unadulterated efficiency.

3. Customer Experiences: Personalized Like Your Morning Coffee

AI's ability to whip up personalized customer interactions is like your favorite barista knowing just how you like your latte. No more spending hours on segmentation and targeting. It's marketing made easy, saving you time and money. And who doesn't love that?

Employee Wellness: Not Just a Buzzword

4. HR Solutions: AI-Powered and Human-Approved

From hiring to high-fiving, AI streamlines HR like never before. It's not just about cutting corners; it's about creating a better work environment. More time for yoga, hobbies, or just binge-watching cat videos? Yes, please!

5. Learning and Development: Tailor-Made

AI's personalized learning paths are like having a private tutor for every employee. It's growth at your own pace, without the time crunch. Think of it as professional development without the professional meltdown.

But Wait, There's a Catch!

6. Pitfalls and Ethical Speed Bumps

AI's promise to "buy time" is as tempting as a midnight snack, but it's not all smooth sailing. Data privacy, ethical use, biases – these are the potholes on the road to AI success. So buckle up and approach with caution.

The Bottom Line

Time's a precious commodity, and AI's like a golden ticket to buy it back. From automating the mundane to fostering well-being, it's a game-changer. But remember, with great power comes great responsibility.

At Fraction AI Consulting, we're all about helping SMEs unlock this potential. Want to focus on what truly matters, like time well spent with loved ones or finally mastering that sourdough recipe? Let's make it happen. After all, time's ticking!

You've mastered your business, but have you unlocked the potential of AI yet? Dive into Aideations - the one-stop-shop for all things AI. With a treasure trove of tools, cutting-edge research, and engaging insights, it's tailored for business savants like you. I've seen the future, blending my Fortune 500 experience with the power of AI. And I bring it to your inbox, with a touch of humor and a whole lot of passion. So, if you've enjoyed the Aideations experience, why keep it a secret? Share it with friends, family, and colleagues. After all, the future is brighter when we build it together.

DISCLAIMER: None of this is financial advice. This newsletter is strictly educational and is not investment advice or a solicitation to buy or sell any assets or to make any financial decisions. Please be careful and do your own research.