I had a chance to look at Kevin's webinar RE: Make automation - honestly, I can think of an easier intro to Make than what he presented. For those of you 'following along' I wanted to highlight that for 'Idea' - don't feel obligated to stick to a keyword - if you published a blog or article use the URL for 'Idea' column - and write the prompt as the video describes - you can create social media posts to go with that article. *And, there's no reason why you can't map that image description to an AirTable column as well in case you want to tweak the image. You can still gen the image w/ Replicate as shown in video but this way you can also gen image using another tool. I took the liberty of extending the example GPT prompt a bit for those of you who do not like risking garbled text in an image: You are a helpful social media manager and content writer that writes platform specific social media content for LinkedIn, Instagram, Twitter and Facebook. Create a realistic image description to go with these posts. The image description must specify that image does not contain any text, letters or words Output in JSON using this format: { "linkedin":"linkedPostGoesHere", "instagram":"instagramPostGoesHere", "facebook":"facebookPostGoesHere", "twitter":"twitterPostGoesHere", "Image":"ImageDescriptionGoesHere" } Also, he mentioned it, but since it was beyond the scope of webinar did not go over Primary Key (PK) aka the Record_ID. I am highlighting this because I've seen a few instances where members of the community have indicated they want to create their own CRM - personally, I think there are plenty of affordable CRM platforms out there but if you feel adventurous - PKs will matter (a lot) along with FKs (foreign keys). So to fully leverage AirTable or other DIY CRM solutions, Google 'normalize a database' and map your entities...

Hey everyone! Ready to grow your social media in the month of December? Let’s commit to this 30 day storytelling plan. Who’s up for the challenge? Copy and paste these prompts into ChatGPT with your LLM training document (AI Business Manifesto) and create a 30 Day Storytelling Content Plan WEEK 1 — Origin Stories Day 1: “I started [business or brand] because I was tired of [struggle your ideal follower relates to].” Day 2: “The moment I realized I couldn’t keep living like that…” Day 3: “If you’ve ever felt [emotion your audience feels], you’ll understand why I created [offer or product].” Day 4: “I didn’t plan to start a business — it happened after [specific event or wake up moment].” Day 5: “This one experience changed the entire direction of my life and business.” Day 6: “I thought I wanted [goal], but what I actually needed was [truth you discovered].” Day 7: “Funny enough, [unexpected moment] was the reason I finally said ‘okay, I’m doing this.’” WEEK 2 — Lessons Learned Day 8: “I used to think [belief], but the truth is [new belief].” Day 9: “Here’s the mistake I kept making that was secretly ruining my [growth or confidence].” Day 10: “If I could go back, I’d tell my younger self this about [topic].” Day 11: “Nobody talks about how hard it actually is to [task related to your niche].” Day 12: “This lesson took me years to learn, but it changed everything.” Day 13: “When I finally stopped [old habit], that’s when things started working.” Day 14: “The day I realized I was doing it all wrong was the day my results started changing.” WEEK 3 — Social Proof Stories Day 15: “When [name] came to me, they were [describe struggle]. Now they’re [result].” Day 16: “The message that made my whole week: ‘[quote or win].’” Day 17: “Here’s what happened when one of my ___ tried [your method].” Day 18: “This result still gives me chills every time [share transformation].” Day 19: “Before using ___, they felt [pain point]. After, they [new feeling or outcome].”

I'm working on a client project, building a very sophisticated web portal/app that has to do a lot of things, with API's, RAG databases, chatbots etc.I started in Lovable but ran into some limitations and last night I told Manus everything the app needed to do, and asked it which of four vibe-coding tools it would recommend for this particular project, and why, and as part of its reply, it produced an interesting table, which I've attached as a screenshot. As you'll see, surprise, surprise, it recommended itself. :) It also seemed to prefer Mocha slightly over Lovable. I didn't include Replit because I'm not fond of Replit since they "improved" it. So I decided to try it in Manus. Amusingly, at first Manus told me it would take 5 people and gave me an estimate of 2-3 months to build it! I asked if this was based on humans doing it, and it confirmed that it was so I asked, "Well, what would it be in Manus time, with just you and me building it?" And it changed its estimate to 6-10 days of work - but it pointed out that because of the speed with which it works, an "8-hour day" to Manus would go by in 5-10 minutes for me, in terms of it actually being done in front of me, and with us working together. Well I worked with it for a couple of hours last night and another hour today, and most of the app is built, looks great, and is working beautifully! The only downside is that so far it has taken about 2,800 credits. I think going forward I'll probably prototype in Lovable or Mocha which use a lot less credits, figure out all the kinks, and then build the real apps in Manus. Manus did some incredibly complex stuff without breaking a sweat, and it also constantly checked its work and when it found an issue, it just fixed it. Overall I'm very impressed with Manus as a vibe coder.

Something I didn't cover last night is that if you scroll down on the main home page of Manus (the one that comes up when you select "New task"), you'll see a lot of user-generated uses of Manus, categorized by different subjects, and what they are is actually replays. So when you click on one, it will replay what Manus did and its end result. It's very useful for ideas and seeing how Manus actually executes them. You can also get to them directly from this page https://manus.im/usecase-from-user I also checked on the credits, and found that creating a Nano Banana style edited image with Manus only uses 13 credits. A VEO 3 video for 8 seconds costs 600 credits, and it says that for 5 seconds it's only 60 credits. But in a test that I did it was actually 108 credits, which is still pretty good. And when you put your video prompt in Manus, it will enhance your prompt pretty significantly to make it more ideal for VEO 3, and you can modify it as you wish before the generation. Here's the link where you can get an extra 500 credits: https://manus.im/invitation/W3A2QNOKWVDOD

My Friend needed a website for her upcoming fitness retreat. Mocha Made this in 10 minutes. It looks SO good. https://adzz5cuvadhjg.mocha.app/. Has anyone else tried Mocha, Kimi or Mixboard from our training last week? Show us what you're up to!