McKinsey: 13 Frontier Technologies, and the AI That Absorbs Everything

A three-family reading framework

McKinsey is now on its fifth edition of this exercise, authored by Lareina Yee, Michael Chui, Roger Roberts, and Sven Smit. The methodology hasn't changed: each trend gets an innovation score (patents and research publications), an interest score (press and web searches), an estimate of equity investment, and an adoption rating from 1 (frontier innovation) to 5 (fully deployed).

This time, the 13 trends are arranged into three blocks. The "AI revolution" groups artificial intelligence and agentic AI. "Compute and connectivity frontiers" bring together application-specific semiconductors, advanced connectivity, cloud and edge, immersive reality, digital trust and cybersecurity, and quantum technologies. "Next-frontier engineering" covers robotics, mobility, bioengineering, space, and energy and sustainability.

This breakdown isn't cosmetic: it shows how radically the profiles differ. AI is a general-purpose technology, rated 4 on adoption, deploying everywhere. Quantum remains rated 1 — potentially transformative for cryptography and materials science, but still far from real business impact despite announcements from sector giants.

AI is no longer a trend, it's a multiplier

The report's most important conceptual shift fits in one sentence: AI is presented as "a foundational amplifier of the other trends." Its impact increasingly runs through combinations — speeding up robot training, advancing scientific discovery in bioengineering, optimizing energy systems.

Methodological consequence: McKinsey drops the distinction between applied AI and generative AI, analyzed separately in previous years, because market solutions now blend both. The numbers follow: interest score of 0.91 out of 1 in 2024, innovation at 0.89, and $124.3 billion in investment, up from $64 billion in 2020.

But the firm is careful to temper this. Realizing AI's full potential will still require innovations to master compute intensity, reduce deployment costs, and finance infrastructure — plus serious approaches to safety, governance, and workforce adaptation.

The two new entries: agents and custom silicon

Agentic AI combines the generality of foundation models with the ability to act: agents that plan and execute complex tasks, communicate with each other, and adapt to new information. McKinsey notes the emergence of both generalist agent platforms and specialized agents, particularly for deep research.

The figures remain modest in absolute terms: $1.1 billion in investment in 2024, an interest score of 0.0033, and an innovation score still rounded to 0. It's the momentum that stands out: +985% in job postings between 2023 and 2024, and adoption already rated 2.

Application-specific semiconductors are the other novelty, and they stem directly from AI. Facing exponential demand for compute, memory, and networking for training and inference — and the need to manage cost, heat, and power consumption — patents have surged. $7.5 billion in investment in 2024, up from $3 billion in 2020, and an innovation score of 0.52, among the highest in the panel.

Where the money actually goes

After a 2023 marked by weak markets and declining funding, the investment climate stabilized and then rebounded in 2024. Ten out of thirteen trends are up.

The ranking is telling: energy and sustainability technologies dominate with $223.2 billion, ahead of mobility ($131.6 billion) and AI ($124.3 billion). Next come cloud and edge ($80.8 billion), cybersecurity ($77.8 billion, more than double 2020's level), and bioengineering ($57.3 billion, sharply down from $127 billion in 2020).

At the bottom of the table, the amounts become almost trivial compared to the media buzz: $9.3 billion for space, $7 billion for robotics, $6 billion for immersive reality — down from $10 billion in 2020 — and just $2 billion for quantum.

In other words, capital remains massively concentrated on energy, mobility, and AI. The most discussed trends aren't always the best funded, and the reverse is also true.

The weak signal: jobs decline almost everywhere

This is the data point the executive summary doesn't highlight. Of the thirteen trends, eleven show a decline or near-stagnation in job postings between 2023 and 2024.

Bioengineering at -17%, quantum at -15%, advanced connectivity at -14%, immersive reality at -11%, space at -9%, energy at -6%, robotics at -2%. Cloud and edge stagnate at +2%, cybersecurity at +7%, mobility at +6%. Only AI (+35%) and agentic AI (+985%) truly pull ahead.

The contrast with the investment rebound deserves attention: money flows toward infrastructure, compute, and physical assets, not toward headcount. And within the tech job market itself, the concentration on AI is visibly happening at the expense of everything else.

Six cross-cutting tensions

McKinsey identifies six themes running through the trends. Autonomous systems — physical robots and digital agents — are leaving the pilot stage for real applications, learning, adapting, and collaborating. Human-machine collaboration models are evolving toward more natural, multimodal, adaptive interfaces, shifting the narrative from replacement to augmentation.

Scaling becomes the real bottleneck. Power supply constraints on data centers, vulnerabilities in physical networks, but also supply chain delays, labor shortages, and regulatory friction over grid access and permitting. The problem is no longer just technical.

Regional and national competition is intensifying: sovereign infrastructure, local chip manufacturing, funding for quantum labs. For McKinsey, it's as much about reducing exposure to geopolitical risk as about capturing the next wave of value creation.

Finally, scale and specialization are growing simultaneously — massive, energy-hungry data centers on one side, low-power embedded models in phones, cars, and industrial equipment on the other. And responsible innovation is becoming a strategic lever: trust conditions adoption.