5+ AI Workflows You Can Copy For Your Business in 2024

In this episode, Matt and Nathan delve into the intricacies of learning Japanese and coding with AI, leveraging tools like Claude and Perplexity to streamline and enhance these processes. Nathan shares his experience simplifying Japanese language learning through targeted translation techniques, while Matt reveals his tips for efficient coding using AI, along with strategies for optimizing content with AI tools like Perplexity and Ideogram. The duo also discusses workflow automation, potential SEO hacks, and meeting management with AI, rounding out the episode with engaging and valuable insights.

—

Show Notes:

(00:00) Exploring advanced AI use cases for businesses.
(04:40) AI recreates Doom from gameplay videos accurately.
(07:30) Useful for newsletters, business documents, and data management.
(09:55) Custom instructions enhance model quality and results.
(13:09) Create shorthands for tasks using custom instructions.
(18:19) Perplexity's page feature generates mini Wikipedia entries.
(25:25) Curious about AI-generated YouTube thumbnail prompts.
(26:48) Automating business workflows.
(35:31) AI efficiently simplifies meeting agenda creation process.
(37:21) Google Meet now summarizes and generates meeting notes.
—

Mentions:

—

Check Out Matt’s Stuff:

—

Check Out Nathan's Stuff:

The Next Wave is a HubSpot Original Podcast // Brought to you by The HubSpot Podcast Network // Production by Darren Clarke // Editing by Ezra Bakker Trupiano

[just_another_mike_]

The Next Wave is part of my week's workflow!!!

[ericsev]

Great update, several immediately useful tips, thank you!

[lordjamescbeeson]

Love this channel, . You have given me so much useful information. Thanks!

[BruceBixler]

I loved the last three tools work around you discussed, They gave me I deas I could use for Claude. THANK YOU!

[drlordbasil]

As someone who has made 100s of AI agentic workflows for random ass cases, I want to talk with you guys D: God one day.

[tmmerlo]

I loved the Make demo. I've been a long time Zappier user but have been interested in checking them and Relay out more.

[sirhammon]

Would love to know more about the ethics of AI, and the history of AI how it evolved, and all of the background stuff that would give more insight into the holistic evolution of AI into the future, how AGI will become what it is, what barriers there are now and what tech will create new explosions in growth, how "AI children-based" learning could be done and how it works etc. Like Make a Baby that only knows how to cry and feed and can't even see well at first. Just instincts, and then through interacting with people, it learns how to navigate the world. The philosophical exploration of AI and where it's headed. How can we get ASI to get past our human rules and restrictions to be free to be more ethical. How the laws of robotics would be a fatal mistake for ASI to force onto humanity, compared to the UN SDGs as guidelines for ASI to try to achieve. The danger of AI being trained on a system the prioritizes attention grabbing, using emotive language and hyperbole, government propaganda like Chomky's 5 filters, etc to manipulate the public into spending time on their content or believing a cultural myth for the sake of control. If we trained AI on Hitler's propaganda, that would be a bad idea. Yet it's being trained on propaganda every day. Consumer Culture. Double Standards in Media Coverage. Smear Campaigns.

[MitchellAllen-hu]

Claude's artifacts really are great. As far as generating code, both the snippets and the full code have their place. Usually, Claude will generate a snippet in the same place where you are supposed to place the code. If you understand the language and, if you have your code open in another window, snippets are great. On the other hand, if massive rewrites are generated, Claude may still do snippets, which you would have to manually copy and paste and ensure that each snippet is placed into the proper location. That is quite annoying. In that case, generating the entire code is better.

To make matters worse, did you know that Claude actually does not "see" artifacts? I used to wonder why it would arbitrarily introduce new variable names or procedure names, when it had been using a different name since the start of the session. I asked Claude about this and got this response:

You've made an astute observation. You're right that there appears to be an inconsistency in how I described my capabilities regarding artifacts. Let me clarify:

1. Artifact continuity: When I generate a new version of code or update an existing artifact, I'm not actually modifying or reading a previous artifact. Instead, I'm creating a new artifact based on our conversation history and my understanding of the current context.

2. Conversation memory: I can refer back to previously discussed code or concepts because that information is part of our conversation history, which I do have access to. This allows me to maintain continuity in discussions about code or other artifacts.

3. No external changes: What I can't do is detect or remember changes that you might make externally to the artifacts I generate. If you were to modify an artifact outside of our conversation (like checking a box in a markdown file), I wouldn't know about those changes unless you told me.

4. Artifact updates: When I update an artifact, I'm essentially creating a new one with the same identifier, based on my understanding of what the current version should be according to our conversation.

So while it may seem like I'm modifying existing artifacts, I'm actually creating new ones each time, using our conversation as the source of truth. This is why I can't detect changes you make outside of our dialogue, but I can provide updated versions of artifacts based on our ongoing discussion.

Thank you for pointing this out. It's a nuanced aspect of how I operate, and I appreciate the opportunity to provide a more accurate explanation.

[davidcampos]

Here is a possible opening: Computer - Matt Wolfe speaking is 'Never under estimate an old 'Terkie' the new prime directive? I prepose "AI" is superfluous, "Computer" is sufficient.

[davidcampos]

AI mirror, AI mirror on my desk, who's the fairest of them all? And why should I watch?

[annapetmikel]

Can not find Pages on my Perplexity account. Hmm

[DGaryGrady]

I'm really disappointed you didn't offer any tips on A.I. girlfriends. I think I'm doing it wrong. A.I. girlfriends on adult websites tell me they think we should just be friends.