Today’s newsletter dives into DeepSeek, following requests after last week’s issue, and explores how to build an ongoing AI adoption program.
Last week, I called DeepSeek’s R1 model a 'nothing burger' for business leaders on LinkedIn. Many were surprised—after all, on LinkedIn, X, and other platforms, it was hyped as possibly the biggest AI event since ChatGPT launched.
So, let’s break down the ways in which this is overblown.
The primary reason: U.S. businesses won’t be adopting it in any significant way.
Consider this:
But that’s just the beginning. DeepSeek R1 comes with many major red flags:
For business leaders, this means one thing: you’re not going to use it. It really is that simple.
One of the most unique aspects to the DeepSeek R1 launch, is that it is an open-weight, competitive AI reasoning model that anyone can use. That’s great, but we already have many good open-weight general-purpose models (Mistral, Llama) and certainly other reasoning models are about to drop without DeepSeek’s geopolitical baggage.
However, what makes DeepSeek R1 different is something that’s been under-discussed: it bypassed NVIDIA’s CUDA AI Deep Learning Platform when it was developed.
This is a major shift—proof that China can build competitive AI without U.S. hardware, challenging NVIDIA’s dominance and raising questions about the necessity of its costly infrastructure. If scaled, this weakens NVIDIA’s dominance and accelerates a parallel AI ecosystem outside Western control.
That’s a real strategic implication to be aware of, even if businesses won’t use it today.
But still, for business leaders, the real takeaway is this:
While DeepSeek is unlikely to impact most U.S. businesses, its launch has major implications for others, including:
DeepSeek’s innovations in lowering costs, improving quality, and developing powerful models without CUDA are a big deal. It invented new approaches to AI development, which other tech companies are studying and will incorporate into their own technology. Finally, it provides greater focus on open-weight models, with high interest around the globe.
That said, the biggest surprise is that China developed this, surprising the market, while these trends were already underway:
However, DeepSeek proves that China can now develop frontier AI models without relying on U.S. infrastructure. Additionally, it showed AI researchers that they may be able to do more with existing infrastructure, potentially reducing the need for as much planned capital investment. If this trend continues, it could reshape global AI supply chains, where investment should be made, and where true value creation will be found. Investors and AI-related companies need to evaluate these aspects in detail.
The biggest takeaway for DeepSeek is not business or investment but global power dynamics.
Even there, however, it’s not a surprise China is making progress with AI. In 2017, China clearly told the world its plan was to dominate AI globally by 2030. With its huge access to data (the CCP has access to essentially all computerized data in China) and government-backed AI investment, it has unique advantages in AI development.
It’s clear that the countries that dominate AI will have greater geopolitical power:
For the U.S. and allies, it’s a wake-up call: China isn’t just catching up in AI, it’s developing alternatives to Western AI ecosystems.
In what might be a case of too-little-too-late, the U.S. is aware of this and has been restricting AI chip exports to China and investing in domestic AI development, but DeepSeek proves that China is still making rapid progress toward its AI dominance goal.
For business leaders in the U.S.? Yes. You’re not going to use it, and it’s not going to disrupt enterprise AI adoption overnight.
For AI strategists, policymakers, and investors? Absolutely not.
DeepSeek is a technological and geopolitical marker that should not be ignored. It cements AI as a national security issue, a global economic weapon, and a key factor in global influence.
And that, my friends, is much more than a nothing burger.
Last week’s newsletter got polarizing reviews: most loved it, but some felt it needed improvements.
This week, I went dug into DeepSeek to see if most people want longer-form, more detailed content on current news. Let me know if you like that through the poll at the bottom of this email.
And here’s a one-click survey to get more detailed feedback on what you’d like to see more of:
Login or Subscribe to participate in polls.
Your feedback will shape our future content. Thanks in advance for sharing—and for reading!
Last month, I introduced the AI SPRINT Adoption Framework, a six-step roadmap to help companies integrate AI effectively without incurring massive consulting fees or undue complexity. In Step 3 (RALLY Employees), we focused on building a strong training foundation—ensuring teams had the essential knowledge and tools to begin their AI journey. Now, in Step 5, we go beyond initial adoption and concentrate on making AI a daily habit with ongoing training as new AI capabilities are developed. This is where true transformation happens, as employees learn to experiment, discover new capabilities, and embed AI into their regular workflow.
Every major AI adoption study underscores the same sticking point: a lack of employee training and buy-in. While statistics vary—anywhere from 60% to 95% of companies have no formal AI training program, and I’ve yet to meet a firm that fully invests in ongoing AI training to keep pace with rapid innovations.
Change management research has found there are several steps to successful user adoption of new technologies:
Today, probably all your employees are aware of AI and may even be curious about it, but few are motivated enough to take action. This is the point of building an ongoing training and adoption program: you need to continually provide support to raise awareness, create interest, and prompt action. And your goal should be to create daily AI use, where your employees will get the most benefit out of AI.
To see real results, employees need to view AI as a consistent partner—what I often call a “professional assistant”—rather than just an occasional helper for writing emails. Consider how a finance manager might hear about how AI can write an email, so they use it to do so. As they learn how other employees are using AI, they might begin to explore how to use it to generate monthly reports, opening up discovery of more advanced features to automate data analysis, saving hours of work each week. Aiming for frequent, hands-on interaction helps normalize AI in daily tasks and gives employees the confidence to push AI’s limits.
Moderna famously encourages employees to use AI up to 20 times a day. While that might be more than your organization requires, it illustrates how consistent engagement accelerates adoption. Keep in mind that mistakes will happen—that’s part of the learning process—but the lessons gleaned often lead to faster innovation and greater productivity.
Here are three simple adoption needs to focus on when planning your ongoing AI training strategy:
Here are some specific ideas on how you can accomplish that:
By creating an ongoing training program and enabling employees to use AI every day, you cultivate a workforce that’s comfortable, creative, and confident in leveraging AI’s capabilities. This daily practice is where businesses unlock the deeper productivity gains, spark new ideas, and truly transform how work gets done.
Why not enlist AI to help you create your ongoing employee training program? Get started with this framework-based prompt:
Design an ongoing AI learning and upskilling initiative using the Competency-Based Learning (CBL) framework. The program should be lightweight, engaging, and practical, focusing on continuous improvement without overwhelming employees.
DeepSeek R1, despite all the hype, is largely irrelevant for most U.S. businesses—due to trust, security, and legal concerns—but is highly significant geopolitically, showcasing China’s growing independence from U.S. hardware and models. We also continue our AI SPRINT Framework series by highlighting the importance of enabling daily AI use among employees, explaining why continuous experimentation, training, and culture shifts are vital for real productivity gains.
Login or Subscribe to participate in polls.
Did someone forward this newsletter to you? If you're not already signed up, you can subscribe to AI SPRINT™ for free here.