Artificial Intelligence (AI) seems to be evolving at lightspeed. 2024 Strongest Man on Earth (SMoE) Mitchell Hooper recently entertained the scenario: Can AI replace strength coaches?
Hooper asked fitness content creator Will Tennyson’s thoughts: “AI is getting too good, no matter how personal you want to get. It can meet your needs,” Tennyson replied. Hooper agrees. so he let AI program the final three weeks of his prep for the 2025 Arnold Strongman Classic (ASC), scheduled for Feb. 28 through March 1, in Columbus, OH.
But first, Hooper faked firing his human coach, Laurence Shahlaei. “I’m switching to another coach…I’m going to try out AI and see how it goes,” Hooper expressed. Unimpressed, Shahlaei responded, “Do what you got to do.”
[Related: Can an Ultramarathon Swimmer Train Like a Strongman?]
The Experiment
Hooper entered all the ASC event details into an AI system, which generated a comprehensive four-week program as prompted:
- Day 1: Lower body max strength day — Heavy deadlift focus
- Day 2: Overhead press day — Log press, stone press, and dumbbell press
- Day 3: Explosive power day — Jerks from blocks, front squats, and speed pulls
- Day 4: Event-specific training day — Frame carries, carry and drag, and stone-to-shoulder
At first, the regime appeared structured and logical, but Hooper noticed some major flaws. “It seems very event-heavy,” he stated, recognizing the absence of general strength-building workouts.
Hooper’s Analysis
While AI got some things right, like progressions, it lacked the nuanced understanding of fatigue and adaptation. “All in all, this program is okay,” Hooper felt, but after crunching numbers, he noted the following issues:
- Inefficient intensity levels: Deadlifts and log lifts were programmed too light, while stone-to-shoulder sessions were excessively heavy.
- Lack of fatigue management: Not accounting for exhaustion when stacking certain movements.
- Failure to recognize human limitations: Didn’t factor in Hooper’s performance limits.
After three days of training with AI’s plan, Hooper was underwhelmed, “It doesn’t really feel like it’s getting me strong like I need to be. I failed a couple of lifts, I feel like I was overworking some things, and other days were just formalities.”
Hooper acknowledged that limited experience with AI may have contributed to the suboptimal results, saying, “I could probably prompt ChatGPT until it gave me the perfect program.” But at that point, he’d rather Shahlaei do it instead.
Missing Human Element
While AI is promising, Hooper doubts it will replace a human coach’s emotional and psychological support. “When you go through something difficult, nothing is more useful than having a friend or an expert there with you,” he explained.
Hooper continued, “AI can solve the problems you know you have, but AI cannot identify the problems you don’t yet know you have.” Since AI programs workouts via inputs, they may lack the intuition to anticipate or respond to real-time challenges.
The Verdict & Future of AI Coaching
Hooper decided to keep Shahlaei’s human expertise, experience, and adaptability. AI may eventually be capable of crafting near-perfect training programs, but it currently lacks humanity and empathy.
Hooper believes AI will play a significant future role in performance programming, pointing out, “The limitation in AI is not AI itself; the limitation in AI is your ability to use it.”
More Strongman Content
- What’s a Force Meter and How Does Brian Shaw Train With It?
- Brian Shaw Reveals 40-Pound Weight Loss Transformation
- Hafthor Björnsson Deadlifts 470 Kilograms During 2025 Arnold Strongman Classic Prep
Featured image: @mitchellhooper on Instagram
