I'm surprised by how many AI components n8n is adding. I would love to see Wappler add similar components.
Agreed. I'd like to see AI incorporated for help/documentation. Take the LLM and train it on all of the docs and community posts. There are several How To's that require updating (some as far back as 2020), where answers to great questions regarding the How To are found in later and separate community posts.
There are known and yet unresolved bugs that could be easily avoided or resolved if the AI were made aware.
Although this post was more focused on AI as workflow components, using it on documentation would be great as well.
The founder of CrewAI mentions all of his documentation was created using CrewAI.
The problem is without a valid subscription (to whichever provider is selected) you'll burn through free tokens so quickly it is not worth the effort implementing it in to Wappler as usage will be severely limited at best. Makes no sense at all as only a minority will have subscriptions, and even then they may be subscribed to an alternative service than that being implemented...
I'm not sure I understand your thoughts. Are you saying that Wapplers don't want to pay for additional services outside of the monthly cost of Wappler, like the cost of tokens for AI?
If so, I feel differently. I am more than happy to pay for it, with the understanding that users like myself are moving towards tools that provide AI capabilities. Not providing the option means many will look elsewhere for the capability.
I would prefer to see Wappler try to keep up with the latest trends in software development. Companies in nearly every industry, especially larger companies, are interested in AI and how it can improve their business or help reduce costs. I think it will permeate every part of our business and professional lives and become even more common than the modern-day mobile device. It boggles my mind that we had landlines, pay phones, and no internet at one point in our lives. I cannot imagine working without the internet, and most companies will need to adapt to this disruptive change or be left behind (Blockbuster, Xerox, Kodak, Sears, Toys R Us, etc.).
Also, regarding the idea of users using different AI services (Claude, OpenAI, etc.), just as Wappler provides the ability to connect to various cloud service providers, the same can be built for connecting to different AI services with frameworks like LangChain. Tools like n8n and crewai use langchain to help with connecting to multiple LLMs in their platforms.
I'm not disagreeing with you Keith. But the tools are already available so what is the difference if it is built in to Wappler or not it will achieve the same thing? What is the difficulty in opening a browser and using it? The output will be the same?
When version 7 arrives I'd like to see stability and historical bug fixes remedied before A.I is implemented. And then I'd welcome it as a further option and capability within Wappler itself.
That is your choice and that of those who make it, and no one is stopping you.
Indeed and we all have faith that the Team will provide for the most common denominators accordingly if Wappler does select to incorporate any LLM. I think that goes without saying. But it still costs money and requires a subscription.
Maybe a poll on the subject to ascertain feedback on that point is required? A sticky poll not one that falls to the bottom after two days and gets forgotten about...
No doubt at some point it will become a part of Wappler. As long as this is a choice and not a requirement then no problem. Then can select to take it or leave it...
What is it you want Wappler to do with A.I that can't already be done by opening a browser? This is the part I fail to understand? Do you just want a terminal to query, then accept and inject or reject its output? Is there something specific you have in mind?
AI built into my app would allow for all types of enhancements.
- Let's say I'm building a CRM system, and I want the AI to have access to the entire schema of my app and connections to different remote repositories of data (e.g., Google Drive or S3) so users can say, "Send Brandon an email with our datasheet about our software as an attachment and schedule a meeting for tomorrow at 2 p.m." The AI would retrieve the document from our file share, send an email with the attachment, and schedule a meeting in our calendar.
- We could build in AI processes to monitor emails or any other record type and categorize them based on the context in the data.
- Still in a CRM, let's say I want to target all contacts in the system that have shown interest in the past 60 days and create a campaign to drip emails to them over time. So I ask, "Create a campaign for those contacts, generate all emails, and schedule them so one email is sent every month."
- We could go further and have an intermediary step to have it research the contacts via LinkedIn or other social media to understand the Contacts better, so drip emails are more personalized.
Maybe this video will provide a better idea of use cases?
Interesting thoughts. I've watched a lot of content on this type of adoption over the past year. I'll take a look at the one you have shared above, tomorrow. I'm supposed to be heading to the train station at 5.30am and it is now nearly 2.30am. Would be great if A.I could attend my meeting for me rather than spend half the day in train stations!
Always enjoy chatting with you Keith, cheers!
Oh boy, that's not enough sleep for me!
Cheers!
My opinion is you should use n8n in your CRM (through API Actions), not worth waiting for Wappler to add their own AI components
n8n will always do it better, given they have a bigger team and such
For Wappler, I'm particularly interesting in having something like Github Copilot for Wappler
Fair point, and that is the route I'm going. I also like the idea of a code pilot, but I think that may be more complex for the Wappler team to implement. Integrating Langchain and connecting it to other Wappler workflow actions may be easier to develop.
Regardless, I just like raising the topic and discussing with other Wapplers.