- Everyday AI
- Posts
- The one ChatGPT mistake that we‘re all making
The one ChatGPT mistake that we‘re all making
Everyday AI
July 09, 2024
👉 Subscribe Here | 🗣 Hire Us To Speak | 🤝 Partner with Us | 🤖 Grow with GenAI
Outsmart The Future
Sup y'all! 👋
When I say YOU call the shots, I mean it.
We asked you yesterday what you wanted to learn about today, and you answered.
Hope y'all enjoyed today's show.
Oh, we're doing a lil giveaway. Make sure to read to the end of today's newsletter. 👀
✌️
Jordan
Today in Everyday AI
6 minute read
🎙 Daily Podcast Episode: You're making one BIG mistake with ChatGPT. We’ll show you why you’re making this mistake and how to fix it. Give it a listen.
🕵️♂️ Fresh Finds: Ex-Meta scientists create AI protein model, Oracle and Palantir announce integrations and Etsy adds AI item guidelines. Read on for Fresh Finds.
🗞 Byte Sized Daily AI News: New York Times found using OpenAI for headline generation, Two new AI video models are released and OpenAI Startup Fund backs AI healthcare venture. For that and more, read on for Byte Sized News.
🚀 AI In 5: We’re reviewing a new data visualization feature from Poe and seeing how it stacks up against Claude Artifacts. See it here
🧠 Learn & Leveraging AI: We’re explaining the one mistake everyone makes when using an LLM and how to fix it. Keep reading for that!
↩️ Don’t miss out: Did you miss our last newsletter? We talked about Intel, Apple and AMD set to rise, China's new ChatGPT competitor, Canva's AI plans and data visualization in Claude Artifacts. Check it here!
The One ChatGPT Mistake That We‘re All Making ❌
You're making one BIG mistake with ChatGPT.
Actually, we're all making the same mistakes when we use large language models.
We’ve trained THOUSANDS of business leaders live on getting the most out of LLMs.
So we know not only WHY we're all making this mistake, but how to easily fix it as well.
Join the conversation and ask Jordan questions on AI here.
Also on the pod today:
• Current Role of AI and Skills 🤔
• Misuse of AI and the Potential Effects 🙅
• Harnessing AI for Skill Improvement 🤹
It’ll be worth your 36 minutes:
Listen on our site:
Click to listen
Subscribe and listen on your favorite podcast platform
Listen on:
Here’s our favorite AI finds from across the web:
New AI Tool Spotlight – Threado AI provides automated support for customers and teams, enso gives you agency-level services with AI and Fluidworks is a live video AI agent.
AI in Science – Ex-Meta scientists have created a gigantic AI protein design model.
AI in Society – Microsoft and Occidental are signing a deal for carbon credits to help offset energy needed by AI.
AI Partnerships - Oracle and Palantir have announced that Palantir's Foundry and Artificial Intelligence Platform are now certified on Oracle Cloud Infrastructure and generally available.
Business of AI – Etsy is adding AI-generated item guidelines to its seller policy.
Read This – Despite San Francisco’s AI boom, office vacancies in San Fran have reached a new high.
Pop Culture – Nicholas Cage spoke on AI and his fears of being digitally cloned.
1. New York Times vs OpenAI: Times secretly using OpenAI’s tech 😬
In a legal showdown of epic proportions, The New York Times has taken a hardline stance against OpenAI and Microsoft, accusing them of scraping articles without consent or compensation.
Despite this, The New York Times has been internally experimenting with OpenAI's technology to generate headlines and articles, aiming to enhance their newsroom operations. This has sparked questions about the future of journalism in the age of AI.
2. OpenAI Blocks China from AI Models, Microsoft Azure Stays Open 👀
OpenAI's recent decision to block API access for Chinese users, effective from July 9, has stirred discussions within the tech community. In contrast, Microsoft Azure, operating in China through a joint venture, continues to provide access to its AI models for eligible customers. This move reflects the growing complexities of international tech regulations and the impact on global AI services.
3. OpenAI Startup Fund Backs AI Healthcare Venture 🧑⚕️
Huffington Post founder Arianna Huffington and OpenAI CEO Sam Altman are joining forces to launch Thrive AI Health, a new venture aiming to build an AI-powered health assistant. This innovative project seeks to provide personalized advice on sleep, food, fitness, stress management, and overall well-being.
With strategic investors like Walmart co-founder Helen Walton’s Alice L Walton Foundation on board, Thrive AI Health is set to shake up the health tech industry.
4. DreamFlare AI Debuts AI-Generated Video Creations 💭
DreamFlare AI is providing a platform for creators to monetize short-form AI-generated content. Co-founded by former Google employee Josh Liss and filmmaker Rob Bralver, the startup offers innovative tools for professional storytellers to craft captivating videos using third-party AI technology like Runway and Midjourney ElevenLabs.
Despite concerns about AI technology threatening traditional jobs, DreamFlare AI insists on creating a new space for creators to earn revenue and democratize storytelling
5. AI Startup Odyssey Unveils Hollywood-Grade Video Model 🎬
Odyssey, the innovative AI startup founded last year, recently shared exciting details about its first AI product - an AI video model that rivals Hollywood-grade visual effects. With a significant $9 million seed funding round led by Google Ventures, the company is set to redefine video editing with a high level of customization and control.
Co-founder Oliver Cameron emphasized the importance of empowering storytellers by providing them with the ability to fine-tune core layers of visual storytelling using multiple AI models.
Poe Previews - Better than Claude Artifacts?
There’s a new feature that was just released from Poe called Previews and it’s pretty impressive!
We’re showing you how it works, when to use it and how it compares to Claude’s Artifacts feature.
Check out today's AI in 5.
🦾How You Can Leverage:
We’re all kinda lazy. 🥱 It’s human nature.
Let’s be honest.
When you go into ChatGPT, you’re probably looking for the easiest and fastest way for that fancy AI thang to do all of your work so you can get back to watching more episodes of The Bear on Hulu.
(For real though, season 3 kinda stinks right?)
But guess what?
That’s actually the exact WRONG way to use a large language model.
Lemme let you in on a lil secret shorties — ChatGPT is LITERALLY trained to be a helpful assistant.
Assistant.
As in, your AI sidekick that is legit trained to poke holes in your process, improve your own skills, and create something sparkly AF that you have zero shot of creating on your own.
Today, we went kinda philosophical on our show about the biggest ChatGPT mistake that we’re all making.
The tl;dr version: we’re using ChatGPT as an extension of our own skillset which short-changes both ourselves and the output we’re trying to get outta the AI.
Don’t worry.
We’ll fix this right up.
First, you should go give today’s show a watch or listen.
We’ll wait for ya here.
(Taking a sip of Coffee #3. Whoops.)
Ok.
Let’s break it down and go over what you need to know. 👇
1 – Trading efficiency for skill development 🤹️
In the short-term?
This is all gravy baby.
It usually looks like this — you find some crazy complex prompt online and swap in your details where the placeholders are and think, ‘yeah, that’ll do donkey.’
Then you hit enter, copy and paste the overly robotic and vague result, and then delve into your next tasks.
Lolz.
What’s the issue?
You just GAMED the system and your manager Bill is gonna have no clue that the 6-hour work just took you 90 seconds. BADABING Bill!
Here’s the rub — if you keep doing this, day after day, hour after hour, something unexpected’s gonna happen.
That main skillset you’ve got?
The thing that you’ve been paid and promoted for?
That specialist task that you’ve been blindly handing over to your new AI bestie?
You’re gonna forget it. You’re gonna lose your edge.
Try this:
We’ll get into the actionable part here more in points 2 and 3, but skill degradation is a real threat to human homies.
Most of your skillsets aren’t like riding a bike. They take ongoing practice, refinement and sharpening.
This article from the World Economic Forum gives great insights on intricacies of reskilling, while this take from the British Council shows us how continuous learning in the age of AI can combat skill degradation.
2 – Your LLM outputs will struggle 😫
Every study shows the best of humans plus the best of AI gives you the best results.
So what happens when you take a shortcut and just want ChatGPT to FINISH THE DANG PROJECT ALREADY?!?
By not putting in the work and looking for the quickest LLM solution possible, you're capping your output at your current skill level.
It's like asking a Michelin-star chef to make you a PB&J.
Tasty? Sure.
Taking advantage of the capabilities at your disposal? Nah.
Why constantly travel a shortcut to mediocrity?
TBH, you’re prolly gonna spend more time turning that overly vague and robotic output into something usable than you would have spent if you did the whole prompt engineering thang the right way.
Try this:
Don’t believe us? Ask the dishes!
(Sorry, had a few random Disney one-liners on deck today.)
If you want a more beautiful output instead of beastly ones, don’t just take our word for it.
Every single study shows that the BEST of humans plus the BEST of AI is gonna give you the BEST results.
Research by the MIT Center for Collective Intelligence found that "superminds" – groups of humans and AI working together – can outperform either humans or AI working alone on complex problem-solving tasks
We mentioned this on today’s show, but helping people and companies get the most outta AI is literally what we do.
More than 5,500 business leaders have taken our free Prime, Prompt, Polish course.
It’s live and Jordan answers all your questions. (For realz.)
Just reply PPP to this email and we’ll send ya access.
3 – Turn ChatGPT into a consultant for best results 🧑💼️
Oh you’re still here?
Lucky you.
Here’s the fix.
Turn ChatGPT into a consultant. Simple as that.
If you followed the advice from #2 above, you’ll get an inside look at our REFINE-Q process, which shows you EXACTLY how to turn LLMs into the helpful assistants that they were trained to be.
In other words…. Consultants!
We’ll make it even easier for ya!
Try this:
Like we said above, the best of humans plus the best of LLMs will always give you the best results.
OpenAI’s own research showed exactly that. Human trainers outperformed those without AI assistance by 60% when using CriticGPT to review ChatGPT code.
But, getting your company or team to quickly adapt to LLMs is’t always the easiest task.
That’s where we come in. Obvi.
Go read some of these reviews. (Apparently, we really know what we’re doing.)
We’re giving away 3 free consults.
To enter?
All you gotta do is go click REPOST on today’s livestream by 1 pm CST on July 10th. We’ll announce the winners in Thursday’s newsletter!
Now This …
Let us know your thoughts!
Vote to see live results
Have you taken our free ChatGPT Prime Prompt Polish Course? |
Reply