Opinion: AI productivity revolution: A practical framework for businesses

Jarad Johnson

At Mostly Serious, we’ve used every artificial intelligence tool out there. We use ChatGPT, Claude, Gemini, Grok, Copilot, Cursor, Bolt, Superwhisper and nearly every other new tool you’ve heard (or never heard) about. We’ve also built custom solutions for our company and clients, including a sales system that made our process four times faster from 2023 to 2024 and will automate all the nonhuman aspects of that work in 2025. These tools have fundamentally changed how our organization works and are soon to revolutionize our industry and world. I don’t believe there’s any false hype in that statement.

But we didn’t start with dozens of tools and custom solutions. We started with ChatGPT. We built our own training (partially because none existed when ChatGPT launched in November 2022), and we equipped our team with a Slack channel to share what worked and didn’t work.

This approach, which prioritizes starting small and learning incrementally, is where I’ve seen so many organizations get it wrong. Some organizations put their head in the sand. They tell themselves this new technology – which is helping some employees complete tasks 55.8% faster, per a GitHub study, and lowering the skill gap by uplifting junior workers – is just a flash in the pan. Other organizations are going all-in way too fast with custom solutions no one will know how to use, solutions that can’t keep up with the evolving model landscape. It’s no wonder research organization Rand found that 80% of AI projects fail, twice the rate of typical IT projects.

The balance is in the middle, where practical benefits are on the other side of novel discovery without overwhelming employees or investing too much in custom solutions before understanding what you really need. Unlike other tools you’ve implemented in your workplace, which likely came from the top of the organization and included clear directions for use and implementation, the benefits of generative AI are best seen through the use cases your employees bring back to you after they’ve been effectively onboarded to your platform of choice.

We train our clients using a three-level adoption framework. We start by equipping people with basic off-the-shelf AI tools, ensuring they get the most out of the immense power available for $20 per month. Then, as users better understand how AI can accelerate or replace aspects of their daily work, we move organizations up the adoption framework to AI workflows and, eventually, fully autonomous AI agents.

Automation
When first introducing AI tools to your team, start with the basics. Pick one system your teams can rally behind and learn together. Ideally, it’s something they’ve already used but in a more secure environment, such as ChatGPT Enterprise or Microsoft Copilot Enterprise. These tools have a low barrier to entry, but with some basic training, employees can significantly improve their effectiveness, efficiency and quality. And then they’re off to the races. By using these tools daily, they naturally identify new areas of their work to automate, opening the door to effective, tested AI workflows you can build onto in the future.

AI workflow
After teams are using the tools daily, you can implement systems that adapt in real time with minimal human oversight. For example, you might build a sales system that delivers a new custom research report to your team with every lead, directly in their inbox. As the sales rep moves through the process, every new piece of material already has a draft completed by AI based on the context of the opportunity and historical information. Estimates, proposals and follow-ups are all ready for human review, executed by their AI co-worker. It’s not about replacing humans but enabling them to focus on higher-value activities, like being present with the prospect and building strong relationships.

AI agent
The most advanced AI tools complete entire tasks without human oversight. Imagine personal assistant agents managing schedules, handling communications and predicting needs by learning from user behavior and preferences via continuous data analysis. Or systems analyzing medical data (images, patient histories) to diagnose conditions and recommend treatments. The promise of AI agents is to redefine what’s possible by tackling challenges beyond current human capabilities.

The path forward
Research confirms what early adopters have experienced firsthand: organizations with AI-led processes achieve 2.4-times greater productivity and 2.5-times higher revenue growth than their peers, according to a 2024 report by Accenture. The key to success lies in measured progression. Start with basic tools that deliver immediate value. Let your team discover use cases organically. Build workflows around these proven applications. Only then consider more autonomous solutions. This graduated approach minimizes risk while maximizing adoption – turning AI from a bewildering technology into a practical business advantage that grows with your understanding.

Jarad Johnson is the founder and CEO of Mostly Serious LLC. He can be reached at jarad@mostlyserious.io.