The internet is on fire discussing a recent New York Times article about how companies are pouring billions into AI which has yet to pay off. The article references a McKinsey report from earlier this summer which says nearly 80% of companies using AI report no material bottom-line impact. McKinsey is calling this the “gen AI paradox.” (Massive adoption, minimal actual results.)
Their ultimate advice is to wait for AI agents to mature so companies can “reinvent workflows from the ground up.” But what McKinsey, (and most of the internet) is missing, is that the workflow reinvention they’re waiting for is already happening. (It’s just not happening where they’re looking!)
McKinsey gets the problem right
The McKinsey report perfectly diagnoses why IT-deployed AI disappoints:
Horizontal use cases (copilots built into productivity apps, enterprise chatbots, etc.) are widely (but thinly) spread across organizations. While they are legitimately helpful, they’re not fundamentally transformative AI. They’re just the latest & greatest tech applied to the existing app stack which is really no different than building a mobile app 10 years ago or a web app 20 years ago. So yeah, they’re cool and useful but not fundamentally transforming how a company operates or showing up in their bottom-line impact.
The opposite kind of IT-deployed AI, which McKinsey calls “vertical” and could drive actual impact, mostly remains stuck in pilot purgatory, with 90% never making it to production. (McKinsey cites all the same old reasons: too complex, too expensive, too much organizational inertia…)
McKinsey’s suggested solution is AI agents which autonomously execute complex workflows. Their report includes case studies with real results, including a bank using “hybrid digital factories” and a research firm automating data quality with multi-agent systems. Impressive!
That said, the case studies also casually mention these projects required new architectures, new tech stacks, new governance frameworks, fundamental workflow redesigns, and (I assume) more than a few McKinsey consultants.
So while I agree with McKinsey that the future is autonomous agents operating at scale, (eventually), they’re missing what’s right in front of them.
The revolution is already here
While McKinsey is talking about preparing for the agent revolution, your workers are already living it. They’re just not using the agents McKinsey envisions.
Your workers are using ChatGPT, Claude, Gemini, and other consumer-focused AI tools as their agents. As I wrote a few weeks ago, these tools are not simple chatbots, but genuine workflow transformation tools:
- Marketing managers build campaign automation without IT
- Financial analysts create models that would’ve taken weeks in hours
- Product managers turn customer feedback into roadmaps instantly
- Engineers write code in languages they’ve never learned
This isn’t horizontal productivity spread thin or vertical complexity stuck in pilot mode. These are real, function-specific transformations happening worker-by-worker, task-by-task, every day.
Your workers aren’t waiting for IT to hire McKinsey to deploy agents. They’re already hiring and deploying their own.
The real paradox: We’re blocking the transformation that works
The brutal irony is that while companies invest millions in custom agent architectures that might work someday, they’re actively blocking the agent-like tools that work today.
McKinsey’s future of “agents supercharging operational agility” is what your workers are literally trying to do right now with consumer AI tools. But they’re forced into absurd workarounds:
- Screenshotting corporate data to upload to ChatGPT
- Copying & pasting between personal AI and work systems
- Maintaining shadow workflows because IT won’t integrate their tools
- Rebuilding the same automations daily because they can’t save them
McKinsey says agents need to be “deeply aligned with company logic, data flows, and value creation levers.” Your workers agree, as they’re manually and painfully doing this alignment every day.
You don’t need to wait for the agent revolution
McKinsey’s AI agent examples are compelling enough to be interesting:
- A bank automating legacy system modernization
- A research firm achieving 60% productivity gains
- Credit memo automation cutting processing time 30%
But these are all top-down, IT-led initiatives requiring months of planning and millions in investment.
Meanwhile, the conversations around your company’s coffee machine are how individual workers are hacking their way to similar gains with $20/month ChatGPT subscriptions. The transformation McKinsey promises with future agents is being attempted right now by workers with today’s AI. The only thing stopping them is corporate policy.
The path forward is simpler than McKinsey suggests
McKinsey’s prescription to build an “agentic AI mesh” by restructuring workflows and deploying sophisticated agent orchestration isn’t wrong, it’s just looking too far ahead.
The immediate opportunities are around enabling the agent-like capabilities your workers already use:
- Secure the environment & workspace, not the agent. Stop trying to control which AI tools workers use and instead focus on securing where and how they operate.
- Enable, don’t replace. Your workers have already identified which AI helps them. Don’t force them to switch to an inferior “approved” chatbot.
- Connect, don’t block. Give workers’ AI tools secure access to the data and systems they need.
- Scale what works. When a worker creates value with AI, make it shareable, repeatable, and scalable. Turn shadow AI into sanctioned innovation.
The work your employees are doing with AI tools today will be the bridge that gets to the world McKinsey predicts.
The companies that win won’t wait
Your workers aren’t waiting for the perfect agent platform. They’re not waiting for IT to finish its pilot. They’re just using tools like ChatGPT to do their jobs better, and if they accidentally “reimagine” a workflow in the process, then so be it.
The “gen AI paradox” isn’t that AI doesn’t deliver value. It’s that companies are investing in tomorrow’s “maybe” while blocking today’s reality.
The enterprises that solve this will be the ones who recognize that the AI revolution is here today, being led by your workers. Yes, it’s messy. Yes, it’s happening in the shadows. The question is whether you’ll enable this, or keep measuring the wrong thing while your workers route around you.
Read more & connect
Join the conversation and discuss this post on LinkedIn. You can find all my posts on my author page (or via RSS).
My upcoming public talks
- AppManagEvent — Closing Keynote: AI & the Future of Enterprise Apps — Utrecht, Netherlands, Oct 10
- MAICON 2025 — AI at Work: The Employees’ Revolution! — Cleveland, Ohio, Oct 14-16
- SHI Summit Fall 2025 — AI in the workplace: What’s happening in the next 12 months — Somerset, New Jersey, 15-16 Oct
- Microsoft Ignite — Citrix’s view on AI in the workplace — San Francisco, 18-21 Nov
from Citrix Blogs https://ift.tt/MuU6N7q
via IFTTT
No comments:
Post a Comment