Activity
Mon
Wed
Fri
Sun
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
What is this?
Less
More
ClickFlow Automation

84 members • Free

AI Profit Boardroom

3.5k members • $59/m

AI Marketing Insiders

1.9k members • Free

Future Faceless Millionaires

7.3k members • Free

AI Freelancers

4.8k members • Free

RC Agency Owners & Freelancers

7.5k members • Free

the skool CLASSIFIEDS

2.1k members • Free

AI Income Builders Club

170 members • Free

AI Automation Agency Hub

333.7k members • Free

19 contributions to AI Marketing Insiders
Where to start?
I'd like to get some recommendations from this community on where to start with AI and eventually monetize that skill. I'm part of several AI communities, and each is pushing its own agenda, tactics, and tools, which is overwhelming. What I like about this community is the marketing aspect of using AI. So, let's say I'm looking to help other businesses with that and other things that AI can do. What should be my starting point?
3 likes • 22d
Start with one workflow you can measure end to end, not a tool collection. Run it manually first and keep a simple log of the input, output, failures, and fixes for two weeks. Then automate only the repeated steps. For tools, begin with ChatGPT or Claude plus the spreadsheet or CRM you already use; wait on an agent platform until the manual workflow reliably produces the same result.
beware, newbie!
I'm a techie, not a marketer. It looks like that needs to change. Did I come to the right place?
5 likes • 23d
Right place, and being a techie is an unfair advantage here - most marketers can describe the workflow but can't ship it. The gap to close is smaller than it looks: pick one offer and one channel, then learn just enough about who's buying, what problem they'd pay to remove, and what happens in the 48 hours after someone raises their hand. That last part is where most systems leak, and it's pure plumbing - exactly your skillset. My advice: don't start by building a big agent. Start by instrumenting something boring end to end (lead comes in -> reply speed -> booked -> closed) so you can see where it drops. Once you can measure it, the AI part becomes obvious instead of hype.
That Opus 5 dip we talked about on the call
Hi everyone, After yesterday's call I went digging into the Opus 5 thing, because I wanted to know if it was just us. It isn't. The complaints are everywhere, GitHub issues with actual measurements, a thread on Hacker News, and even Anthropic's own status page has an incident open for it. Plenty of people are finding Opus 5 worse than Opus 4.8 in day to day work, even though the benchmarks say it should be fine. So a practical thought rather than a moan. ChatGPT is worth using at the moment to cover the things Opus is fumbling. And if you have both in VS Code you get the best of both worlds anyway, you can run them side by side and pick whichever does the job better on the day. I use more than one model as standard for second opinions on my own work, and weeks like this are exactly why. These dips do tend to get looked at and fixed, so I'd treat it as an adjustment for right now, not a reason to switch camps. Anyone else noticed particular tasks where it's got worse? Would be useful to compare notes.
2 likes • 23d
Same pattern here. The place it bit me was long multi-file refactors and anything needing exact numbers from a doc - it got confidently sloppy. Two things that helped without switching camps: (1) pin the model version in whatever you're running instead of "latest", so a bad week doesn't silently change your output, and (2) keep a tiny regression set - 5-10 real prompts you run with expected outputs - and re-run it when a version ships. Takes 10 minutes and turns "feels worse" into something you can actually show. For marketing copy tasks I've barely seen a difference; it's the reasoning/tool-use chains where the dip shows.
LinkedIn and the use of AI
Hello everyone, I am glad to find such a great community and all the post I've read do far were adding so much value. I have a question on LinkedIn and AI. 1) Has anyone experienced any penalisation when using Claude on Chrome - my idea is to research suitable posts to comment on as a daily skill, but also research suitable connections to my ICP through Sales navigator. The comments and invitations would be done manually. 2) Has anyone experience any penalisations when using scheduling tools ie Blotato. I am not planning to do anything in scale -we are talking maybe 10-15 comments a day and 5- 10 connections a day, but I wouldn't want to make any enemies 😈 Happy to hear how else you are using AI on LinkedIn as I am trying to be more active and grow my network. 🙏
3 likes • 24d
Your proposed split is the safer one: keep AI to read-only research and first drafts, then make the actual connection/comment decision and final edit yourself. Avoid browser-driven bulk actions, vary cadence instead of batch-sending, and keep a simple log so you can stop if acceptance or reach changes. For schedulers, use native or approved integrations where possible and treat third-party posting tools as a separate risk from research. I would test one workflow at a time for a couple of weeks rather than introduce research, invites, and scheduling together.
From ranking on Google to being the source AI names
Welcome to everyone who's joined recently, good to have you here. If you want to just read for a bit before you post anything, that's fine, most of us did that. The good stuff is in the replies though, so jump in whenever. Here's something I've been reading up on. AI is starting to just name a source, rather than give you ten links to wade through. Yelp has licensed its reviews, photos and business data to OpenAI, so ChatGPT can now recommend actual businesses with current ratings. That was late July, and they've since added booking, so you can reserve a table without leaving the chat. The IAB, the trade body for the digital advertising industry, has also put out the first proper standard for measuring how visible your brand is inside AI tools. Measuring Visibility in the AI Era, out on 3 August. So the target has moved. It used to be ranking on Google. Now it's being the thing the AI names when someone asks. One caveat though, and it's a big one. There are already more than twenty companies selling AI visibility tools, and they all measure it differently, so you can get different answers for the same brand. That's the IAB's own finding and it's why they bothered writing the standard. Only about one in six brands is even tracking this yet. So if someone offers you an "AI visibility score" this year, ask how they worked it out before you believe it. Two questions. 1.Has anyone here seen their own content cited or summarised by ChatGPT or Gemini? And are you doing anything deliberate to make that happen, or is it still down to luck? 2. And if you have seen it, did it get you right? I'd like to know whether these things are citing people properly or just roughly.
1 like • 25d
I’d treat this as two separate surfaces: licensed partner data can win recommendation slots where the model needs current structured facts, while earned visibility still matters when it is synthesizing advice. The practical response is to make your first-party facts consistent and machine-readable across the pages you control, then measure citations and referral queries separately from classic rank tracking.
1-10 of 19
Igor Ganapolsky
3
27 points to level up
@igor-ganapolsky-7420
AI missed-call text-back & lead triage for contractors. Get audit: buy.stripe.com/14A4gz9eA8Hkf0lfSH3sI1B?client_reference_id=skool_profile

Active 19h ago
Joined Jul 15, 2026
ISTJ
Fort Lauderdale, Florida
Powered by