[AI SPRINT] 75 orgs. 0 AI plans. Gallup data + new ChatGPT tools

I spoke with about 75 organizations last week.
Not one had a real plan for AI.

That’s not just anecdotal—it’s backed by Gallup’s annual AI adoption report.

• 40% of employees are already using tools like ChatGPT at work.

• 19% use them daily (and as I often share, daily users see ~20% productivity gains).

• Leaders are twice as likely to use AI as their teams—33% vs. 16%.

• But only 22% of organizations say they have a concrete AI strategy.

That last stat is the real story.
In my conversations with business leaders, the top reason is clear: it’s hard to know where to start—and even harder to keep up with how fast the technology is changing.

That’s why I built the AI SPRINT framework—and this newsletter.
But it’s not enough.

This summer, I’m expanding my team and exploring new ways to support you.
I'd love your input.

👉 Take this quick poll—just click the option that fits best!

Last week, OpenAI rolled out new updates to ChatGPT, enhancing its Canvas and Projects features to make it an even more powerful tool for creators, professionals, and teams. These updates streamline workflows, boost collaboration, and simplify how work is exported and organized—reinforcing ChatGPT as a central hub for both creative and structured tasks.

But beyond the features themselves, this reflects a broader shift redefining how we work.

ChatGPT (and other emerging AI systems like Claude, Grok, and Perplexity) are evolving rapidly from conversational assistants into control centers for your entire workday. No longer just “help me write an email” AIs, they are becoming full-fledged operating systems for knowledge work.

Just two weeks ago, I shared how ChatGPT now connects to key enterprise systems like Microsoft Outlook, SharePoint, and virtually any other application via the Model Context Protocol (MCP). This means your AI can now access your data (CRM, ERP, etc.), understand your personal context and preferences (through memory), and act with autonomy (via agents, Operators, deep research, and task scheduling).

With tools like Canvas (for writing), Codex (for coding), and multimodal capabilities like image generation and voice, ChatGPT and other general purpose AIs will become the center of your workday. Even on mobile devices.

We have entered a phase where your primary role shifts from “doing the work” to instructing and supervising AI that does the work for you. These latest updates point clearly in that direction.

Whether you're drafting a report, debugging code, or organizing a complex research project, these new capabilities make ChatGPT a more seamless, integrated, and hands-on assistant—reducing the friction of switching between apps and accelerating how work gets done.

So, make sure you are testing and beginning to use these new capabilities to stay at the forefront of AI adoption! Keep reading for more detail, and a quick look at the environmental impact of ChatGPT.

Overview: What’s New with ChatGPT

1. Canvas Exports

The Canvas content editor now enables you to download content as PDFs, Word documents, Markdown files, or code files (e.g., .py, .js, .sql). This makes it easier to share your work with others and external tools.

2. Projects Enhancements

The Projects capability—designed to organize your work and track progress over time with dedicated knowledge and structured folders—now supports deep research, voice mode, enhanced memory for recalling past chats, and mobile file uploads. These upgrades transform Projects into powerful, persistent workspaces ideal for long-term initiatives. For example, I create Projects for each of my technical projects, and for each client, to track everything in one place.

3. Expanded Platform Support

Canvas is now fully available on web and desktop (Windows and macOS), with viewing and interaction on mobile.

These updates make ChatGPT more collaborative, portable, and tailored for complex workflows—whether you're a solo creator or part of a team. More details below, plus the latest on ChatGPT’s environmental impact.

Canvas Updates: More Than Just an Editor

Canvas, ChatGPT’s dedicated workspace for writing and coding, is evolving from a drafting interface into a multiformat publishing and development environment. You can still add basic formatting (bold, italics), collaborate on sections (like asking to “write in a more professional style, as below”), and approve and review suggestions.

New, is that you can now export content as PDFs, Word documents (.docx), Markdown files, or code files like .py, .js, or .sql. This makes it dramatically easier to move from ideation to execution—sharing, storing, or deploying your output across tools and workflows.

In addition, Canvas now renders HTML and React code, and supports full editing of model responses and code blocks—letting you iterate directly within the workspace without copy-pasting into other tools.

Why It Matters

• Seamless Sharing: Export a report as a polished PDF, a blog post as a .docx for editing, or production-ready code as .py or .js files—right from Canvas.

• Integrated Dev Support: With HTML/React previews and live editing, Canvas supports end-to-end prototyping for web content and applications.

• Snapshot & Versioning: Canvas maintains draft history and allows quick restoration of earlier versions, aiding both solo work and collaboration.

Usage Examples

• Marketers can draft a campaign brief with ChatGPT, refine tone and style in Canvas, then export to Word for final formatting and review.

• Engineers can debug a script collaboratively with ChatGPT, annotate it, and export directly into their IDE.

• Designers can generate and tweak React components or HTML layouts, preview the results live, and export for front-end implementation.

But note: I still find Canvas buggy, particularly for long-form content and conversations. To get the most out of it, use it for tactical content, like editing an email.

Get started by using the Edit in Canvas Pencil icon, or just asking “Open in Canvas”.

Projects Updates: Your Persistent Subject Workspace

Projects are ChatGPT’s solution for organizing work over time, with persistent memory, dedicated instructions, and file organization. Think of them as AI-powered folders with brains. They help keep all your related content organized, and enable you to “chat” with the content directly.

With the latest update, Projects now support:

• Deep Research: You can now use Deep Research within a project to tap into external sources like Google Drive, SharePoint, Dropbox, or GitHub—pulling in relevant files and data for richer analysis.

• Voice Mode: Speak to your Project hands-free for brainstorming, summarizing, or querying content—especially useful on mobile.

• Improved Memory: Projects now better remember previous chats, context, and instructions, providing consistent continuity across sessions.

• Mobile File Uploads: You can now upload files directly from your phone and continue workflows wherever you are.

Why It Matters

• Centralized Workspaces: Each Project keeps your chats, files, goals, and instructions in one place—ideal for long-running tasks or multi-stakeholder work.

• Context That Sticks: ChatGPT retains memory within a Project, so your assistant isn’t starting from scratch each time.

• True Portability: With mobile file upload and voice interaction, your Projects are accessible and actionable anywhere.

• Toolchain Integration: By connecting to services like GitHub or Google Drive, Projects become deeply embedded in your real-world stack.

Usage Examples

• Researchers can collect papers, run deep searches across Google Drive, ask ChatGPT to summarize findings, and brainstorm thesis structure via voice—all in one Project.

• Event Planners can centralize budgets, timelines, and vendor contracts, then request AI-generated action plans or timelines with stakeholder-ready exports.

• Developers can analyze codebases pulled from GitHub, reason through architecture with ChatGPT, and export the updated code when ready.

Environmental Impact of ChatGPT

People always ask me: Is ChatGPT bad for the planet? Now we have a figure to judge it by. In a recent blog post, OpenAI CEO Sam Altman shared that each ChatGPT message consumes approximately 0.34 watt-hours of energy and one-fifteenth of a teaspoon of water.

If we assume a “frequent” user of ChatGPT sends 20 messages a day to ChatGPT, 30 days a month, that’s equivalent to:

• 204 watt-hours of electricity—running a 60-watt lightbulb for 3 hours.

• ~6.7 fluid ounces of water.

Most independent estimates suggest that AI systems consume significantly more energy and water, so these low estimates are encouraging to see!

But there’s a catch. Altman’s estimates don’t include:

• Training Costs: The environmental cost of training large models—often involving thousands of GPUs—was not factored in.

• Energy Source: The impact varies depending on whether electricity is sourced from renewables or fossil fuels.

• Geographic Context: Resource consumption differs by region based on data center location and local infrastructure.

• Community Competition: Whether AI infrastructure diverts water and power from local communities was not addressed.

• Lack of Independent Review: These figures come from a personal blog post, not a peer-reviewed environmental assessment.

While it's a positive step to begin quantifying environmental impact, more transparency and standardized metrics are needed. With AI usage expected to grow dramatically, robust data will be essential to responsibly manage its footprint.

Recommended Next Steps

• Try Exports in Canvas: Start with a small project—export a doc as Word format, or a code snippet as .py, and test how it fits into your real workflow.

• Launch a Project: Pick one regular type of work you do (e.g., research, planning, content dev), create a Project, and upload supporting files. Use deep research or voice mode to explore.

• Experiment with Mobile: Upload a file or use voice mode from your phone—see how AI support travels with you.

• Collaborate with Instructions: Tailor a Project’s instructions for your team or task. ChatGPT will respond with context-aware support.

Which new feature are you using first? Hit reply—I’d love to hear how you’re applying this!

About Trent: Trent Gillespie is an AI Keynote Speaker, CEO of Stellis AI, former Amazon leader, and advisor on building AI-Native, AI-Enabled businesses. Book Trent to speak to your group or book a call to discuss using AI within your business.

Did someone forward this newsletter to you? If you're not already signed up, you can subscribe to AI SPRINT™ for free here.