AI’s latest breakthrough? Humans

It's pretty much impossible to go a day without stumbling across another AI breakthrough. Albania has 'the world's first AI government minister', and AI-generated music is topping the charts. An Oxford study has used AI to help classify the night sky, and scientists in China claim to have developed the world's first 'brain-like' AI. But among industry hype and existential hand-wringing about AI hoovering up jobs, a new market paradigm is quietly shifting into place. The latest AI frontier isn't artificial general intelligence, robot ministers, or digital brains; it's humans. 

As AI weaves into everyday lives and companies attempt more experimental use cases, it might seem like humans are playing second fiddle. But as models get increasingly complex, humans are actually taking on a new prestige. Cue the latest offering in AI: 'humans-as-a-service'. 

So, why are humans the hottest thing in the AI market? It's because the world has rapidly realized what we've known all along: to function at their best, AI tools need experts in the driving seat. 

AI needs its engineers

On a fundamental level, engineered systems across all levels of complexity rely on human guidance to remain sharp. With AI, this means that human oversight remains crucial for training models, correcting inaccuracies, and repairing systems that need attention. If it's designed and engineered, it will always need mechanics.

This principle scales with complexity. Consider something as simple as a bicycle: punctures, loose chains, and broken brakes all require specialist tools and know-how to properly repair. A  $100,000 racing bike demands far more specialized attention than training wheels. AI follows the same logic: the more sophisticated the system, the more expert intervention it needs. 

This becomes clear with real-world applications. Higher-value use cases like digital twins aren't just harder to set up; they demand specialized maintenance and have complex failure modes that need experienced practitioners to navigate effectively. 

For businesses, this creates a challenge and an opportunity. The opportunity lies in increased productivity, improved accuracy, and the ability to generate previously unseen insights. The challenge, however, is making sure these systems are underpinned by strong data governance foundations. No matter how advanced your AI capabilities are, if your people don't have access to the information they need, they won't be able to integrate this data into AI applications or workflows. 

This challenge scales when considering that few teams currently have the spare resources, digital skills, or in-house talent to balance their own responsibilities alongside getting data ready for AI systems and troubleshooting technical issues. AI itself can automate these processes and fix data quality issues, but human oversight and guidance remain critical for deciding where AI can help most, and how to prepare data to support AI use cases.

As systems become more powerful seemingly everyday (with AI 'superintelligence' on the horizon), maintaining this software – and keeping it on track to deliver value – exclusively in-house becomes less and less realistic. Human experts will only become more and more important for successfully onboarding, maintaining, and leveraging AI in business.

The Consulting Boom 

Unlike most AI predictions, this shift towards 'humans-as-a-service' isn't future-gazing; it's happening now. This summer, OpenAI invested $10 million in a brand new AI consulting arm, while in September AI consultancy firm Synechron reached a $1 billion valuation. The market is recognizing that successful enterprise AI implementation depends on strategic human guidance. 

This is because AI projects with no clear direction don't deliver value as a quick fix. A recent MIT report revealed that a staggering 95% of enterprise generative AI projects fail to deliver the rapid growth companies expect from them, citing poor integration and reliance on generic tools as key factors. 

Stapling a chatbot like Claude or Gemini onto your current work setup might deliver some value for productivity. But just because an LLM can produce lines of code in an instant, it doesn't mean it can improve your processes and help you hit your KPIs. Instead of providing low-value generic utility, AI in the enterprise is most impactful when it delivers an answer to a well-defined problem. 

This is where expert squads of 'humans-as-a-service' are truly invaluable. Humans excel at understanding nuanced situations, specific business needs, and contextual AI use cases that can seamlessly integrate into existing workflows. Deciding on these use cases is where humans can add significant value. Expert consultants can advise on the best off-the-shelf solutions for your needs, or develop tailored AI solutions to real problems and save employees' time that can be better used for more strategic tasks.

Whether this looks like using AI to streamline routine processes, automate data governance, or provide intelligent search capabilities, working alongside a human with the expert tools and knowledge will produce an AI roadmap that is realistic, scalable, and has measurable impact. It's about finding the right tool for the job, and keeping it going. 

As artificial intelligence grows increasingly sophisticated, human intelligence becomes increasingly integral for successfully delivering return on AI investment. The irony is, on the surface, perfect. But this shouldn't surprise us. As AI's master craftsmen, humans aren't becoming obsolete – they're becoming more important than ever.