30 Ideas for an Applied AI Working Group
Initiatives for your company-wide AI innovation team
Random Observation/Comment #816: I like to use “please” and “thank you” in my AI prompts. All hail the glow cloud.
//Generated on MidJourney. I find it hilarious an AI overlord bot would just be typing away on a computer with explosions in the background.
Why this List?
I believe in the decentralized approach to learning and implementing applied AI for personal productivity and company strategic differentiation. It would be super cool to be the “Chief AI Officer” (CAIO), but a bottoms up approach for this technology is more impactful. This list of 30 will hopefully give companies and those key leaders some ideas on initiatives.
Company-wide Policies – Your legal and security team should have done this back in the ChatGPT Spring frenzy in March/April 2023. The main one being the confidentiality agreements on sharing data with external tools. You may also want to let your teams know if they should be creating burner accounts or connecting their Google Suite emails to these tools and services.
Dedicated #applied-ai slack channel – We started this in April 2023 and encouraged everyone to share relevant news articles and discussions. It’s still a lively place to read up on pivotal material and try out new tools
Create an #applied-ai notion page – Have a standing page where people can get started with science experiments, recordings, key mentors, and resources.
Provide access to tools like Azure’s openAI playground – We had a large number of Azure credits so we worked with their team to get company-wide GPT3.5 access for the hackathon
Embed AI into your Hackathon – We hosted a session on how to use prompts and examples of early day implementations. One of the hackathon awards highlighted the use of AI.
Host monthly Applied-AI sessions – I’ve been running these 1-hour sessions since May 2023 with a lot of engagement and curiosity. My strategy was to distill the latest news and continue to refine the thesis around how Web3 could be impacted and improved
Create standing training sessions on prompting or for specific roles – It’s incredibly important to just use the tools available on a daily basis. Keep the page open and consider prompting and fine tuning instead of Googling.
Research what your competitors are doing – In the Web3 space, I’ve found most external-facing tools to be either “Chat with gitbook docs”, “ChatGPT Plugin for “, or “Chat with customer support discord bot trained on our material”
Write a weekly internal newsletter – Get people excited about the technology by providing a few reading and watching links with a mini assignment
Build an internal-facing private ChatGPT bot – There was a leak in early august of an Azure ChatGPT – https://web.archive.org/web/20230814000219/https://github.com/microsoft/azurechatgpt – I’m sure a private “ChatGPT for Business” is coming soon.
Train your marketing team for external-facing material – Embed this into your content writing and editing process. Consider summarizing and rewriting in a different voice your competitor articles. Create content at scale. Consider creating material like shorts and tutorials from longer videos.
Test out public facing bots – I’ve tried botpress and AI-Engine wordpress plugin and they’re both launched on the bottom right side of seelemons.com
Talk with partners and tools about their AI pivots – See how they’re enabling employees and consider similar policies
Host a dedicated day for employees to explore AI – Yay Prod Web3 Days!
Explore product impact and features – Brainstorm embedded features or new angles on user experience. How will FAQs be different?
Consider writing pieces about using AI/ChatGPT – Tutorials and articles for leveraging AI are getting much higher SEO results
Add AI talks to your adjacent programs – Consensys added talks about AI to the Ambassador Program, NAVH Hackathon sessions, and DAO conversations
Launch a team – These can probably be passionate volunteers. I imagine a younger group of new joiners would contribute to this with more energy. I don’t think you need a full-on “AI Innovation Lab”, but that all depends on your budget.
Write a prompt database – Map certain deliverables to specific prompts that can provide context and thorough results. Share optimized prompts.
Start a “AI Tools Review” – During the session, demo how specific AI tools are making the creator economy much more efficient
Join AI Summits and Hackathons – This is always a great way to meet other passionate people and see upcoming technologies. Conference go-er could be a full time job.
AI Awareness Campaigns – Deploy an internal campaign with some catchy names and cool graphics to get people immersed in the technology
Show and Tell – Ask for people to share how they’ve been using AI in their daily operations. Do they use it mainly for brainstorming and writing? How have some outputs been augmented by AI?
Guest speakers – Look for prominent speakers from summits to host a session directly with your team members
Resources List – Provide access to Books, Podcasts, and Courses for getting up to speed with the latest technology
Adding AI to OKRs and performance goals – This could be a bit extreme, but it could be motivating if there’s ways to look at optimization and tooling
Ask how current vendors are leveraging AI – The IT tools you’re using are likely going to have some AI improvements. This could be Notion AI for writing, Figma Jam chatbot, Duet AI in Google Docs and sheets, Canva AI image generation, or Zoom AI for meeting summaries. There’s tons of new features coming out that will be embedded in your existing workflow.
Employee recognition – It doesn’t hurt to point out people who have been vocal or active in moving this initiative forward
Communication – Definitely be clear how the company is approaching AI. Some categories could be: Education, Evangelism, Internal Productivity Tooling, Internal Infrastructure, Product Impact, and Experimentation
Ask ChatGPT how to start an internal program – Always Be Prompting. While brainstorming is my superpower, so is using the tools of the trade.
~See Lemons Apply AI to Consensys
Originally posted on seelemons.com