These are the 7 basic yet overarching principles for any employee or organization to become AI-native that I shared at an introductory AI adoption workshop at AppsForBharat that our team at Elevation conducted yesterday. Big thanks to founder Prashant, Tanvi Lal and Trisha Bahri for enabling this.
> Default to AI Mindset: As a rule of thumb, assume AI can do everything. AI should be the default thought when blocked or needing assistance. Don't hesitate - just try using AI for any task.
> Start with pain points: Look for moments in your workflow where you think "oh man, I just can't do this, it's boring". That exact point is where you should explore automation. Don't try to automate everything at once - solve specific pain points first.
> AI is personal always: AI works better when it's true to your personal workflows. Build automations that match how you specifically work. Create your own personal prompts and context files.
> Model-Format Fit is a thing: Don't try to do everything with one model (like ChatGPT). This is the biggest reason most think 'AI is not great for me yet'. Find the right tool for each specific task rather than forcing any one to do everything. Not every model is designed to do every task optimally. Find yours.
> Pair programmer approach for org adoption: Engineering teams know how to automate but don't know what to automate. Non-engineering departments know what to automate but not how. The solution is every department should have an engineer on standby to help automate. Else you're ngmi. Now is the time for all silos in an org to work closely together.
> Open library culture: Don't gatekeep AI knowledge. Proactively share learnings among teams. Create shared prompt libraries and system prompts across the company. They'll come in handy for someone at some point. The distinctions between roles (PM, designer, developer, sales) have blurred with AI. Help the analyst learn how to sell and the salesperson how to make products.
> Continuous exploration: You need to constantly explore new tools. If you find something remotely interesting, sign up and try it. Take up free trials. Shoot some prompts. If you're not satisfied with the results, repeat after a month. The model or tool would've improved. You must be aware when the right tool for your niche launches anywhere.