“All the technology in the world will never replace a positive attitude.”

—Harvey MacKay, Author

Bridging the GenAI Divide

MIT’s latest State of AI in Business report shows that most companies are stuck. While more than 80% of organizations have piloted AI tools, only 5% of those efforts reach production.

Despite over $40B invested, 95% of GenAI projects fail to deliver measurable ROI.

The problem, they’ve discovered, is not a lack of adoption. It’s that most tools don’t learn, adapt, or fit the workflows they are meant to support. And no one wants to use a tool that doesn’t work.

What’s Actually Working
While official deployments are stalling, 90% of employees are already using personal AI tools like ChatGPT for real work. At the same time, only 40% of companies provide official access.

Over 50% of GenAI budgets are going to sales and marketing (on things like outbound emails and lead scoring) the highest returns are showing up in back-office automation.

The most savvy businesses are using it to replace expensive vendors and accelerate manual work. In service, finance, and document-heavy processes, they are eliminating millions in BPO contracts, cutting creative and agency spending by nearly a third, and saving more than a million annually through automated risk and compliance checks. These wins are showing up fast, and without major changes to team structures.

What Business Can Do
The data points to a practical path forward. Leaders who want to get real value from AI are:

• Auditing how their teams are already using AI.

• Considering all departments. ROI is often highest ignored functions like operations and finance.

• Choosing vendor tools that integrate easily with existing systems.

• Shifting attention from front-end features to back-end efficiency.

Sometimes you need your AI to be a coach, not a cheerleader. If you're trying to pressure-test a strategy, get honest feedback, or spot blind spots in your thinking, try this prompt setup:

Most LLMs are trained to be agreeable and emotionally supportive by default. That can be helpful in some situations, but if you're making tough decisions on things like tech, hiring, or process changes you don’t need emotional padding. You need a second brain. This prompt cuts through the simulated warmth and invites the AI to challenge your thinking in a useful way.

• Management & Operations: Maskara
  An AI platform that helps users create custom AI agents and design more effective prompts using just a few user inputs. It simplifies automation by letting you deploy tailored AI tools for tasks like moderation, summarization, and personalized recommendations without needing deep technical knowledge.
   

• Marketing & Advertising: Ocoya
  A social media assistant that brings content creation, scheduling, and performance tracking into one platform. It streamlines everything, from writing and design to publishing and analytics, so marketers and business teams can manage posts efficiently across multiple channels.
  

• Content Creation: Writesonic
  Helps teams generate marketing copy, blog posts, and product descriptions with built-in tools for SEO and publishing. Designed to streamline content creation from writing to distribution.

In our latest episode of Auto Collabs, Laurie Halter brings real numbers and real dealer feedback on how AI is showing up in service.

She joins Paul, Kyle, and Michael to talk about which tools are earning a place in the shop, why predictive maintenance still feels out of reach, and how smarter systems could help close the technician gap that everyone is feeling.

Catch the replay here.

• Sam Altman responded to GPT-5 backlash by updating features and restoring GPT-4o. ⏪

• In just 90 days, over 100M AI videos have been created with Google’s Veo3. 🎥

• Rumor has it Meta will reveal their “Hypernova” smart glasses next month. 🤓