Friday, August 15, 2025
Read time: 5-6 min
Read this article online
Hi, it’s Chad. Every Friday, I serve as your AI guide to help you navigate a rapidly evolving landscape, discern signals from noise and transform cutting-edge insights into practical leadership wisdom. Here’s what you need to know:
1. Algorithmic Musings. The Three Leadership Lessons Hidden in OpenAI’s Rollout Mess
Last week I promised you a rundown on the new GPT-5. I’m glad we waited, because if you’ve followed ChatGPT during this period, you’ve likely felt like you were riding a technological roller coaster while blindfolded.
The saga began when OpenAI decided to channel their inner Thanos and eliminate all legacy models with one swift snap, replacing them with GPT-5. Users didn’t exactly throw a parade. The rollout was immediately followed by backpedaling, mea culpas, and a hasty rollback that would make a politician blush.
The dust has finally settled. Plus and Pro users can now select one of three “modes” for GPT-5: “Auto,” “Fast,” and “Thinking.” Free users? They’re stuck with an icon that simply says “ChatGPT.” Think of it like being handed a menu with only one item.
What Actually Changed (The Tactical Stuff)
Most users will stick with “Auto,” but power users now have real control options. ChatGPT Plus users now get 3,000 messages per week with GPT-5 “Thinking.” GPT-4o has been restored by default for paid users, because apparently some things are worth keeping around.
Plus users (and above) now have access to several legacy models: GPT-4.1, o3, and o4-mini. You’ll find this tucked away under Profile > Settings > General > Show Additional Models. This advice is fairly tactical, yet this feature is buried deep enough that many folks will never locate it without guidance.
Other meaningful enhancements include:
Multi-session project continuity: Your projects now persist across multiple conversations instead of vanishing every time you open a new chat. This means greater efficiency and consistency across separate drafts.
Expanded multi-modal capabilities: Significantly enhanced ability to process text, image, audio, and video inputs. Think seamless capture of a whiteboard brainstorming session and its conversion into a polished text document.
Increased context window: GPT-5 can now handle up to 400,000 tokens in chat, enabling the system to tackle larger documents and more complex workflows.
The Not-So-Great News
No roundup would be complete without acknowledging the problems:
Fabrications: Despite my disdain for the term “hallucinations” its usage persists. That said, while the system improvements are meaningful, there’s significant chatter that the level of outright fabrication has also increased. I have not seen reliable data demonstrating this to be true but keep in mind that accuracy and innovation don’t always hold hands so please check your homework.
Not-so-smart model router: GPT-5’s strength and occasionally its weakness is its back-end model switching based on judgment calls about the task at hand. This works beautifully… except when it doesn’t.
Broken workflows: If you’ve built custom GPTs and workflows on older models, they likely experienced significant disruption. Consider this a cautionary tale: AI companies may no longer provide advance notice of model deprecation. You’ve been warned.
Three Leadership Lessons From This Digital Train Wreck
I wouldn’t be doing my job unless I called out the leadership lessons hiding in this technological mess.
1. Force Kills Adoption (Even When You’re Right)
Users didn’t ask for GPT-5 to replace their familiar GPT-4. Yet this mistake gets repeated constantly in organizations everywhere. How often do we impose “improvements” on our teams without considering their attachment to current processes?
OpenAI learned that even objectively better technology can face revolt when deployed without proper preparation. Excellence without adoption equals failure.
2. Quick Backpedaling Beats Stubborn Pride
When the user backlash hit, OpenAI didn’t double down or explain why users were wrong. They listened, apologized, and rolled back. How many leaders dig in their heels when facing similar resistance?
Ego is expensive. The fastest way to turn a small problem into a catastrophic one is to defend a bad decision instead of fixing it.
3. Choice Restores Control (And Buy-In)
The solution? Give users three modes instead of one mandate. “Auto,” “Fast,” and “Thinking” transformed a forced march into a menu of options.
This approach works because users are often paralyzed when faced with too many options. Three, though? That’s just enough choice to feel empowered without triggering decision fatigue.
What This Means for Your Next Big Change
Before you roll out that new system, reorganize those departments, or implement that strategy, ask yourself these questions:
- Are you forcing a change or offering options?
- Have you prepared your people for what’s coming?
- When resistance appears, will you listen or lecture?
Change is hard enough when people choose it. When you impose it without warning, you’re not just implementing a new system. You’re declaring war on the status quo. And the status quo has more allies than you think.
The Bottom Line
OpenAI’s rollout mess offers critical lessons in how not to implement change, followed by a redemption story about how to fix it. The difference between disaster and success often comes down to three simple concepts: preparation, humility, and choice.
Got a major change on the horizon? The OpenAI playbook might just save you from your own rollout disaster. Drop me a line if you’d like to talk through your change strategy. I promise not to eliminate your legacy processes without warning.
2. Sound Waves: Podcast Highlights
Our latest episode features Evan Schwartz, Chief Innovation Officer at AMCs Group and Forbes Technology Council member. Evan reveals how middle-market fleet operators are missing massive efficiency gains hiding in plain sight – including how saving just 2 gallons per day per truck generates $700 daily for a 100-vehicle operation. Key takeaway for executives: The ROI on fleet AI isn’t in the flashy stuff. It’s in the multiplication effect of small, consistent improvements across route optimization and predictive maintenance. Subscribe for free today on your listening platform of choice to ensure you never miss a beat.
New episodes release every two weeks.
3. Research Roundup: What the Data Tells Us
AI CREATIVE WRITING: THE CONTEXT SOLUTION TO BLAND CONTENT
New research just answered a question I get constantly: “How can we create unique-sounding content with AI?”
The numbers that matter: When researchers gave AI models just 30% of a story as context, creative output diversity matched human authors across all measured dimensions. Even five random words in prompts improved variety almost as much as detailed context.
What this means for your Monday morning: Stop blaming AI for generic content when your team is feeding it generic prompts. Your marketing copy can maintain distinct voice—you just need to give AI something to work with.
Your next move: Audit your AI content workflows this week. If your prompts are under 50 words, you’re creating the blandness you’re complaining about.
Vendor reality check: AI still skews 70% positive in sentiment versus humans’ balanced emotional range. Factor this into brand voice guidelines.
Read our full analysis of this and all other analyzed research papers at AI for the C Suite.
4. Radar Hits: What’s Worth Your Attention
Perplexity AI bids $34.5 billion for Google’s Chrome browser. While this deal is unlikely to close, it signals how AI companies are positioning for control of internet infrastructure. Translation for executives: the future of search and browsing is being rewritten by AI-first companies. If your digital marketing strategy relies heavily on Google search dominance, start diversifying your customer acquisition channels now.
AI traffic is up 527% and SEO is being rewritten. This isn’t speculation – it’s actual traffic data from 19 companies showing ChatGPT, Perplexity, and other AI platforms are becoming legitimate discovery channels. Legal, finance, health, and SMB services are seeing the biggest impact because people ask AI consultative questions they’d normally ask an expert. The takeaway: if your content strategy still only targets Google, you’re missing a fast-growing traffic source that your competitors are already capturing.
5. Elevate Your Leadership with AI for the C Suite
Three C suite execs called me this week asking the same question: “How do we implement AI without creating our own rollout disaster?” If you’re sitting on an AI initiative that’s been stuck in committee, let’s talk strategy. I’ll help you avoid the OpenAI playbook of forcing change and dealing with revolt later. (Shameless plug: I promise not to eliminate your legacy processes without warning.)
As we navigate this unprecedented fusion of human and machine intelligence, remember: the best leaders aren’t just adapting to change – they’re actively shaping it. Until next week, keep pushing boundaries.
Chad