Why GPTs can't drive executive strategy + raising the bar for humans

The crash and burn of AI tooling in practice - here's what we know

My love-hate relationship with AI is complicated - as a technologist with 25 years in digital - but one thing I’m 100% confident about is that:

GPTs will never drive true strategic roadmaps

Running a successful business in 2025 is contingent on different factors, including:

• Developing a product or service in need

• Achieving product-market fit with niche features or solutions

• Building a strong audience that trusts you

• Tapping into a data MOAT of self-developed studies and analysis that’s not public knowledge

• Strategic partnerships with other industry leaders in the space

• Forward-looking perspective and understanding of macro dynamics and the ecosystem (industry insights and predictions)

• Effective operational efficiency under the hood to deliver the corresponding results

Realistically, AI in 2025 is only able to support the last bullet point here - optimizing internal processes in large organizations with thousands of people doing mundane tasks, conducting manual research, scraping data by hand.

A few years ago, automation for most of that was already possible - with custom Python scripts, Selenium bots, headless Chromium scrapers, and other systems. It wasn’t necessarily trivial to develop them + some maintenance was required, which is why virtual assistants or admin roles were filling that gap without breaking the bank.

But none of the content production workflows, or deep fake AI videos, or automated email outreach at scale is able to build meaningful connections, retain professional communities, and craft scalable products that last.

And while a random influencer with an active community and 3 million followers can prove you wrong, they can achieve the same with a $10K prototype build, a Bubble app, an outsourced agency content flow for $3K - $5K per month, or any other high ROI medium to speed up development and go-to-market for a product that’s already signed off.

GPTs can upskill your low-skill talent, or mid-level managers lacking basic industry know-how. It may bridge the gap for non-fluent speakers able to respond in fluent English or even reply in a different language. But these are light enhancements for operational processes, and not driving true go-to-market strategy.

From a technical standpoint, this could go wrong real fast. I shared the story of this entrepreneur launching a public product with paid subscriptions and getting hacked via trivial endpoints:

AI has been known to provide fake legal and medical advice, get basic arithmetic wrong, and parse spreadsheets with hallucinations as well.

And in terms of business strategy…

All of that holds true because GPTs are being fed public data sources

I challenge agents and LLMs weekly on core business challenges around:

• GTM motion mechanics that work

• B2B audience development

• Lead generation for 2025

• Sales and TAM expansion

• B2B influencer marketing plans

• Scoping out new business plans (i.e. best businesses to invest $100K for $5K/mo with a breakdown)

• Topics, resources, and tools for our audience of B2B leaders and executives

Here’s a sample deep research that took Gemini nearly 10min to run - and what data sources were used for the analysis:

Scraping Google for regurgitated and generic data

But unless you’re launching a commoditized agency for TikTok videos, an influencer marketplace - or a good old dropshipping store (like the tens of millions of other Shopify ones), then copy-pasting a beaten-to-death plan is what you want to run away from.

The real value in business today is kept in:

📰 This is an excerpt of the original issue published on Wednesdays. Read the full article + join the email list of 16,000+ executives enrolled in this newsletter here 📰

📲 Want to work 1:1 with me or hire DevriX as your enterprise partner? Here's how.