Justin's AI Newsletter

The most interesting content, news, and insights in AI each week.

--

Hey friends 👋,

Greetings from Colorado!

We've been off to the races getting more companies onboarded onto my product, Cassidy.

My team and I have been personally onboarding every customer. Although it's not the most scalable strategy, I strongly believe in investing time early on to truly understand our customers, the problems we solve, and ensure a fantastic experience for everyone.

If you're on the waitlist, we'll reach out soon!

---

AI STRATEGY

💬 Leveraging AI to Supercharge Content Creation

The Prompting Technique to actually produce high-quality content.

I really believe that every business in 2023 needs to have a solid content strategy. This means producing hundreds of high quality pieces of content across various platforms weekly. Thanks to AI, its never been easier to build a powerful content strategy.

Now, let’s be clear: a strong content strategy does not mean using ChatGPT to flood the internet with hundreds of generic blog posts vaguely related to your business. In fact, I think the vast majority of people can already tell if something was written entirely by ChatGPT, and will mentally write off the content.

So, whats the secret to making your AI assisted content sound genuine rather than robotic? A technique known as Few-Shot Learning.

Few shot learning is a technique in prompt engineering that allows you to recreate your authentic writing style with LLMs.

Instead of having AI just write your content for you, you need to provide examples of the ideal output. This ensures that the content you receive is on-brand and high-quality.

Let’s go through a specific example.Imagine you’re a skincare brand trying to write Instagram captions. Without Few-Shot Prompting, you might prompt ChatGPT with:

​

Now, the result is fine. However, it lacks knowledge of our ideal writing style or our unique business specifications, and it just sounds generic.

Here is the prompt with few-shot learning:

​

Few-shot learning involves providing examples of an input, and using those examples to teach ChatGPT what it should return. Here's a video tutorial.

My suggestion for scaling your content production is building a library of "Few-Shot" Content Prompts. Create unique prompts for each content type, whether it's LinkedIn posts, TikTok scripts, or Threads.

When you have this library of working prompts, you can go from half-baked content ideas to high-quality posts in minutes.

Integrating AI-Enabled Content Strategy Across the Team

Few-shot prompting is one of the many prompt techniques that power a feature called “Writing Templates” in my product, Cassidy.​

This feature focuses on solving this exact problem: creating content that is authentic and relevant to your organization incredibly quickly. We combine the power of few-shot prompting with your company’s internal knowledge and data.

If you want to create a library of writing prompts that go beyond ChatGPT and share them with your entire team, sign up for Cassidy’s waitlist.

This is something we’ve worked on extensively, and companies have been using it to rapidly produce many pieces of content a day.

(I’m giving early access to my newsletter subscribers…)

---

💡 This Week's Recommendations

🔗 Code Interpreter available to everyone: Everyone is getting access to the Code Interpreter this week. It's a powerful tool that I highly recommend you play around with if you have ChatGPT Plus. It's still early, but it shows the power of AI in the data world. Check out this thread of examples.

🔗 Business leaders fret about generative AI despite growing enterprise adoption: Study: Many Fortune 500 business leaders are planning to adopt generative AI for productivity gains. However, they are also concerned about issues such as quality, safety, and security. They expect AI to change data-driven roles and assist in research, but they have concerns about job displacement, reduced innovation, budget constraints, and legal compliance. Ultimately, I expect AI adoption to occur in every Fortune 500 company over the next 5 years.

🔗 How to Make AI UX Your Moat: This article makes a great point that UX is the crucial factor. When AI startups are criticized for merely being a "wrapper on GPT," it's evident that's what ChatGPT was. I've said it before, and I'll say it again: the reason ChatGPT went viral is that the chat UX unlocked the possibilities of LLMs for the masses. I believe innovative UX work will continue to be the differentiator for AI startups.

🔗 OpenAI introduced Superalignment team: OpenAI has formed a new team, Superalignment, to tackle the challenge of superintelligence alignment. Basically, 20% of their compute power of the next 4 years will be focused on using AI systems to evaluate other AI systems, automate the search for problematic behaviors, and detect misalignments. This just goes to show how significant of a problem alignment is.

---

Thanks for tuning in,

Justin

​

P.S. I want this newsletter to be super useful for you—so I'd love to hear your thoughts! What do you want to see more or less of? I'll reply to everyone :)

​Justin's Full Content Library | Previous Newsletters​

113 Cherry St #92768, Seattle, WA 98104-2205

If you'd like to update your email settings, choose one of the options below.

1) Unsubscribe from my weekly newsletter.

2) Unsubscribe to be removed from all future mailings, including occasional and highly important AI updates you won't want to miss. That'll make me sad. But hey... I get it. You're busy. Just know that once you click this link you won't receive any more emails from me.

Other Preferences​