AutoScripting 90 Days in 90 Minutes

Full video transcript

How’s it going everyone? Darby here with the AI Author, and happy Thursday, happy Thursday to everyone out there, and hope your New Year is off to a great start. I’m excited to be here with you today because over the course of the next 90 minutes to two hours or so I’m gonna be going over a recipe that I’ve created in collaboration with Jon Benson called the Buyer’s Blueprint recipe. This is going to be the 3.0 Edition. Now I worked with Jon Benson on this recipe starting probably about five, six months ago now, and this reps what we got out of this recipe, and creating it was the reason behind the recipe: first and foremost was to create the different variables and functions inside Jasper, brainstorming with Jasper to create the variables for the CopyPro marketing blueprint. That basically uses Jasper’s superpower brainstorming and doing customer research and market research.

As you’ll see inside of this recipe, that’s the heavy lifting of creating this marketing blueprint, and once you put all of these outputs inside of Jasper, inside of CopyPro, now you have all of these really proven direct-response email funnels, sales pages, and landing pages centered around a very specific avatar. Who are you solving a very specific problem for? They have a goal in mind, and so all of your marketing copy is super coherent. Really powerful combination when you’re using the two of these tools together. But today I’m going to be going over how I’ve been using the same exact recipe that you guys are going to be able to access if you’re a part of our AI Author community. This recipe will be available inside of our Gen AI Insider newsletter, which I’m going to be wrapping up the final bits of after this live stream and sending out later this evening or tomorrow morning. So this recipe, along with this recording so you can watch and learn how to use the recipe for content creation, it’s gonna have a lot of elements in it, and talk of prompt engineering right now is really hot with ChatGPT and all of that, so we’re going to be using ChatGPT as well, because I honestly haven’t run the prompts from this recipe inside ChatGPT yet, so we’ll be able to compare the two side by side and see how these commands and prompts compare. My experience with ChatGPT so far is that it’s good for certain things. I find that sometimes it can rewrite emails in different tones of voice, or rewrite things in a less aggressive, more passive tone depending on how you’re directing it, and I find that sometimes it does that better than Jasper. But I’m still experimenting with it, and with little creative stories here and there both work side by side as of right now. Of course ChatGPT is free, Jasper is paid once you’re really starting to use it. My personal issue with relying on ChatGPT right now is that there’s so much volume that they’re receiving that their servers have been down half the time whenever I go to the site. I’m not relying on it professionally, but I’m experimenting with it. I do find there are places it shines in certain cases, but I’m still Team Jasper for sure, and we’ll see how things go — obviously this industry is changing very fast.

Before we kick things off here I want to say hey to everybody joining us live — let me know where you guys are tuning in from. I have a few ideas for different projects right now that we’re going to need to create a lot of content for, so I’m going to be experimenting and brainstorming which one — probably going to go with our UC Berkeley Change Makers course for AI and education, because we’re ramping up for the semester right now and kicking off our course end of January, early February, for open enrollment. So we’re going to be creating a lot of content over the next 90 days about AI and education: its implementations, different ethics around it, problems with it, and content that speaks to this industry and how AI is changing things. We’ll see what Jasper can help us with, because I’m really excited about this course. The thing with a lot of content out there teaching you how to do stuff is that it’s great to watch and see it, but with a tool like Jasper or ChatGPT you can literally follow along right now, so you can walk away from this workshop hopefully having 90 days’ worth of content that you’ve created for your brand, keeping more content in front of more customers.

Before we kick this off, if you want access to this recipe, go to theaiauthor.com key links — I’ll show you exactly where this recipe is, along with a whole bunch of other recipe goodies. Once you get here you’ll notice there are a lot of these recipes; a lot of these are in our cookbook as well, and there are key links down here if you want the cookbook, a new recipe template if you’re into making recipes, and if you haven’t taken your first steps with Jasper, go through the Jasper Boot Camp first — that’s the number-one place to start. A few other things down here, like our AutoScripting micro-course, if you haven’t gone through that’s how you get signed up. These are all free recipes for you guys to use, and they’ve gotten a lot of really good reviews. For this recipe, go over to the CopyPro Jasper Commands tab, and up here, F1, the Buyer’s Blueprint Recipe 3.0. Click that link and you’ll come to the Buyer’s Blueprint Recipe 3.0. There’s a lot of training we have on actually implementing it into the marketing blueprint inside CopyPro, but we’re not going to be doing that today. What I want you to do is go in here, grab that recipe, and click Run. Once you click Run it’s going to open up the recipe — it’s a big recipe, over 3,000 words inside of it right now — and now the recipe will be inside your Jasper account. I’m going to come out so I can show you exactly where this recipe is in case you want to go back and reference it: go back to your documents, and from there go down to the Recipes section, right below Documents on the left-hand sidebar. In here you’ll find community top recipes and the recipes you’ve created. I’d recommend bookmarking the key links tab too, just so you have it on hand.

As we get started with the recipe, I’m not joking when I say John and I spent dozens of hours on Zoom breaking down all of these different commands, and right now if you go into CopyPro, most (not all) of these commands are integrated inside its own tab so you can copy and paste them in. This is a more advanced approach, using recipes to get things going. We’re about 15 minutes in; I’ll have about 15 minutes at the end if we time this to 90 minutes properly, but I’m not going to be prepping this content right now — my focus is to get the content out and batch the work for this project, and after this I’ll hand it off to my team, who will start prepping it, turning it into graphics, Twitter posts, and more over the next 90 days.

As we get started with the recipe, step one is filling out your content brief, because this content brief is key for Jasper’s context. Context is everything, and this is really where you want to engineer your prompt to have the most effective outputs on the commands (I call them commands, not prompts, because they’re building off certain levels of context and used in specific situations). The content brief holds all the variables inside your prompt that you’ll need — avatar, niche, and more — but you don’t need to fill out all of these at every stage. In fact, at certain stages in this recipe I’m only going to start with two, plus editing the title, which is how Jasper references context, so I set the title to “AI and education [Buyer’s] blueprint.” My avatar in this case is Educators, and my niche is Education. Background here: the AI Author’s coaching business is working with Kelly Tenney, a 20-year veteran in the education space, to create a course with UC Berkeley’s Berkeley Change Makers program, starting this spring semester — officially announced just last week. So the avatar is teaching high-school and college students through an eight-week course, but the blueprint is built around Educators as the audience, since they’ll see different perspectives on AI after the content is generated. I’ll work with Kelly to turn round one of the content into round two, then prep it for post-production.

With avatar and niche set, you don’t strictly need the primary complaint to start (though it’s good to know what your prospect’s number-one concern is), but the next most important variable to add is the primary goal, because it feeds all of the secondary goals the recipe generates. I set the tone of voice to visceral, emotional language, which tends to add more color and better outputs in testing, and this also matters because CopyPro formats content in specific ways for compliance and other reasons. As an example (one Jon Benson filled out previously): avatar = women 50+ in the weight-loss niche, primary complaint = “they can’t lose weight,” primary goal = “losing weight.” Once you have avatar, niche, primary complaint, and primary goal, Jasper can generate all the other variables.

Beyond the content brief, you’re at the “begin recipe” stage, where you start brainstorming with Jasper. You, the human, remain the ultimate decider on whether an output is good enough and where to go next — you’re very much brainstorming, and some outputs will be great, others not so good. Running a command is as simple as clicking Compose (Ctrl/Cmd+J), or using Cmd/Ctrl+Shift+Enter, or the “run next sentence” button in the upper right. Cmd+Shift+Enter behaves a bit differently than Cmd+J — worth experimenting with both, since results vary. For the primary-goal command, Jasper generated: “The primary goal of Educators is to inspire and motivate their students to learn, grow, and reach their full potential…” Darby ran it a few more times to get multiple outputs, organizing them with headers so they can be pasted into a Google Doc afterward. He noted that Jasper sometimes “prompts itself” into related directions (like secondary goals) that are still useful, since much of this content later feeds into building new recipes.

Moving into topic generation: the recipe generates lists (e.g., “10 examples”) that can be used later; you don’t need to be picky yet since not everything generated will get used. A live question came up about whether deleting a rejected Jasper output counts against your monthly usage — Darby wasn’t fully certain but noted that on the unlimited plan he doesn’t see the effect directly; he suggested using the thumbs-down / “poor completion” flag instead of just deleting bad outputs, since that’s more likely to get credits refunded.

Darby then paused to run the same primary-goal command in ChatGPT for comparison, using the identical prompt and context. ChatGPT returned similar phrasing (“helping students reach their full potential,” “empowering students to reach their full potential”), which Darby found comparable to Jasper’s output on this particular prompt, noting you’d want to keep ChatGPT and Jasper prompts organized separately since he’d lost content in ChatGPT’s interface before without a clean way to organize it.

For the primary complaint, Jasper generated a 10-item list such as “feeling overwhelmed with all the information they need to know in order to teach effectively,” “struggling to engage students,” and “spending too much time trying to develop lesson plans.” Darby explained that each of these list items is a potential subject line, blog post topic, LinkedIn article, or Twitter thread. For the secondary complaint, the recipe generates roughly 25 examples; Darby explained (crediting Jon Benson) that secondary complaints matter because “what you’re selling is on all the secondary complaints that people have,” and addressing those secondary complaints is what actually converts people who have the primary complaint. He then updated the content brief to add the primary complaint itself, noting that outputs became noticeably more specific and on-topic (referencing AI directly, rather than generic teaching content) once that variable was filled in — illustrating that Jasper doesn’t infer intent, it needs it stated explicitly in the brief.

A viewer (Carl) asked whether recipe creation is covered in Darby’s book; Darby pointed to the AI Author key-links spreadsheet, the free cookbook, and the AutoScripting micro-course at autoscripting.ai as the places to learn recipe-building in more depth. The micro-course, roughly two hours across nine sections, has been running since around September of the prior year and covers what Darby called the three pillars of AutoScripting: prompt engineering, command stacking, and workflow automation with Jasper recipes. He noted the course is the backbone of an upcoming book, “AI and The Rebirth of Writing,” planned for release on Amazon and Kindle. The free AI Author Cookbook (at theaiauthor.com) walks through the full Jasper Boot Camp and organizes recipes by chapter, e.g., “need FAQs? Click the link and the recipe loads directly into your Jasper account.”

Returning to the recipe, Darby reviewed the 25 generated topic ideas, noting the outputs became “way more in line” once the primary complaint was added to the brief versus the earlier, more generic pass. He then demonstrated Power Mode, which lets you generate content directly alongside a working document (rather than a separate recipe view), picking one topic — “fear of being left behind as AI-based education becomes more mainstream” — and generating a one-shot blog post. In parallel he ran the same prompt in ChatGPT (“write a compelling blog that educates and entertains the reader on the topic of…”) to compare outputs. Jasper produced a blog with a title, introduction, and body (“Keeping Up with AI-Based Education: Don’t Get Left Behind…”); ChatGPT’s version covered similar ground but Darby noted ChatGPT returned a network error mid-session, reiterating his point about relying on Jasper for reliability. A viewer asked what Darby does with blog copy once generated: he said he typically edits it himself inside Jasper docs for shorter pieces, or hands it to his team for further editing; for books he uses a separate tool (Living Writer) to organize chapters and sections pulled from Jasper output.

Continuing down the recipe: the objections section generated items like “I don’t want to rely on AI to educate our students” and “AI sacrifices the personal touch,” which Darby framed as a problem/solution pairing pattern borrowed from Alex Hormozi’s “$100M Offers,” useful for addressing objections directly in content. The ultimate fear section generated consequences like “risk of dropping out of school and losing motivation” and “missing opportunities to develop critical skills and knowledge.” Darby took one of these into ChatGPT to generate a Twitter thread using “visceral emotional language,” comparing multiple outputs and picking the one he found most usable, then pasting it back into Jasper’s document to keep everything centralized.

He also demonstrated Jasper syntax nuances — for example, that Cmd+Shift+Enter has roughly a 150–200 character limit per run versus Compose, and that adding a colon after a command primes Jasper to continue directly. The promises section generated reasons AI could benefit education even without a specific “product” filled in. The mistaken beliefs section produces paired “mistaken belief” / “the truth is” outputs designed for CopyPro’s blueprint fields, which Darby also repurposed into a Twitter thread format (“There’s a lot of misconceptions about what AI means in the classroom… here are 10 hard truths about AI when it comes to education”).

Moving through “the enemy” section (framing an antagonist/cause), Darby noted that adding an explicit “1.” to a prompt increases the odds Jasper returns a listicle format rather than a paragraph. He worked through consequences, negative statistics, and secondary goals sections, generating items like “increasing student engagement,” “developing critical thinking skills,” and “nurturing digital literacy.” He explained the “false solutions” and “success myth” sections, and a command-stacking pattern — where a first command generates a list (e.g., physical/emotional/psychological/spiritual challenges) and a second command builds directly on that list’s outputs (numbered C1 through C9) to generate solutions, which then feed into Twitter thread prompts. He noted that small wording tweaks in a command stack can meaningfully change results, and that command stacks generally only need to be run once when they work correctly.

Near the end of the recipe, Darby ran the offer-reframe section, generating benefit-framing copy like “It’s not just an AI course, it’s a way for students to become empowered change-makers… It’s not just education, it’s a way to make a meaningful change in the world with AI technology,” directly usable for the UC Berkeley course marketing. He finished with an offer-target command stack, generating a list of audience segments (career changers, tech professionals transitioning into teaching, entrepreneurs) and corresponding offer copy for each.

By the end of the roughly 90-minute session, Darby noted the recipe started at about 3,000 words of prompts and produced about 12,000 words of generated content. He said the outputs are directly usable inputs for CopyPro’s marketing blueprint (to build full email sequences, sales pages, and funnels) as well as raw material for social content. He thanked viewers for joining, invited questions and feedback by email, and mentioned a related five-day paid workshop (“Trio AI”) starting January 16 with Jon Benson, Jasper affiliate partner manager Sean Vosler, and other copywriting and marketing experts, focused on using CopyPro, Jasper, and ChatGPT together to build email sequences and copywriting businesses (limited to 100 seats). He also pointed viewers to the Gen AI Insider newsletter (a low-cost weekly curation of AI industry news, prompts, and community access) and the Cookbook Plus community for deeper, ongoing training. He closed by reflecting on how Jasper had helped him overcome writer’s block and “prompt paralysis” in his own content creation, and signed off until the next live stream.