The Next Great Reallocation of Work
What do you get when you cross the human need to work, with robots and AI?
For most of human history, work has been inseparable from survival. If you wanted to eat, stay warm, or belong to a community, you worked. Technology has steadily reshaped how that work is done and we are probably entering another one of those moments.
The current wave of AI and automation is beginning to reallocate work across the economy. Not all at once, and not everywhere, but in ways that are already visible. AI code tools can reason, generate, and coordinate. Robots can increasingly operate in semi-structured physical environments. The predictions tend to swing between abundance and collapse, however we think both extremes miss the point.
This is not primarily a story about jobs disappearing. It is a story about how people move through work when technology reshapes roles, closes some doors and opens others.
We think AI will materially reshape work, but unevenly. Some sectors will absorb change quietly, whilst others will feel it sharply. The biggest risk is probably not mass unemployment in the near term, but a widening friction: people unable to signal their skills, to navigate change, or to access the next set of opportunities as roles evolve.
This is an issue not just for economic stability (unemployment -> low productivity -> increased fiscal spending… etc.), but also for quality of life (jobs -> livelihoods -> fulfilment and purpose). And this raises an important question - is this moment any different to the major tech-driven transitions we’ve gone through in the past?
This deep-dive is an attempt to understand the opportunities this presents - the ways to transform how we manage careers, jobs and hiring. We look at previous technological shifts, what’s different now and the areas we’re excited about investing in.
From this work we think three areas stand-out as underdeveloped:
tools that enable genuine skills-based hiring at scale;
systems that help people navigate careers, reskill, and transition repeatedly; and
infrastructure that supports a growing class of independent workers and entrepreneurs as the traditional employment contract weakens.
If you’re operating in these spaces, whether currently ideating, or have founded your company, we want to hear from you. Reach out on albert@revent.vc.
Keep scrolling for the full deep-dive
A brief history of tech-driven workforce transitions
Technological change rarely hits labour markets all at once. More often, it arrives unevenly - early adopters, then the followers, some geographies before others, and so on. Three examples illustrate this pattern.
Example 1: Automotive manufacturing and absorption rather than collapse
Automation in automotive manufacturing is often cited as evidence that robots destroy jobs, but the data tells a more nuanced story.
Since the early 1990s, US vehicle output has increased by around 8 percent, while employment in motor vehicle and parts manufacturing has fallen by only around 3 percent. This sits against a broader backdrop in which total US manufacturing employment has declined by roughly one third since its 1979 peak.
That wider decline was driven not only by automation, but by offshoring, trade liberalisation, and global supply chain reorganisation. Auto manufacturing is structurally different. It is capital intensive, politically protected, and relatively anchored to domestic production. Productivity gains were absorbed through higher output per worker and changes in task composition rather than large-scale labour shedding.
The lesson is not that automation has no effect, but that its labour impact is shaped by many confounders, such as sector economics, labour institutions, and exposure to global markets. It also shows that technology can give individuals leverage, not necessarily replace them.
Example 2: Agriculture and the long unwinding of farm labour
In the early twentieth century, a large share of the workforce in advanced economies was employed in agriculture. Mechanisation, fertilisers, hybrid seeds, and better logistics transformed output over decades, not years.
The results have been quite dramatic - the current state of agriculture is markedly different from Agri in the 80s and 90s. Agricultural output rose severalfold while employment collapsed. But the collapse was slow, stretched over generations, and buffered by migration, industrialisation, and eventually the expansion of services. The chart below shows real farm output in the US and employment intensity of that output (both series set to 100 in 1948). This clearly shows the heavy labour impacts of technological development in this sector.
This remains one of the cleanest examples of technology directly reducing labour demand at sector level. It also shows how long these transitions can take.
Examples 3: E-commerce and the lag between adoption and reorganisation
E-commerce is a great study in an ‘internet’ driven job market impact. It also shows some of the nuance around hype cycles for new technologies. When things started to be sold online, many were calling for the end of the high-street, however even today we still only see around 15% of goods sold online.
The chart below shows US data mapping the employment intensity of retail sales (left axis), with the rising Ecomm share of retail (right axis). Despite this somewhat slow and steady ramp, the absolute employment impacts have been very significant.
Noisy, then slowly, then all at once?
Many new technologies share a common pattern. They arrive noisily, early demonstrations impress, but deployment is patchy and whilst productivity gains are real, they are localised. In effect, labour markets ‘absorb’ the change.
Then, often quietly, the conditions for scale can fall into place. Costs drop, workflows adapt and complementary infrastructure emerges. At this point, transitions that once seemed unlikely, start to fall into place. Anecdotally, we see three broad trajectories for job market impacts:
We think looking at these patterns matters because much of today’s debate assumes a smooth, linear adjustment, but history suggests otherwise. In the three examples in the previous section we see a variety of responses. Manufacturing was noisy, then slowly, then all of a sudden production was largely fully automated. Agriculture was slowly, then steady. E-commerce has been noisy but the absolute impacts have been relatively flat.
We think the challenge at stake is that we underestimate how quickly jobs in sectors can shift once adoption passes key thresholds, and therefore how poorly prepared individuals and jobs markets are to support that shift.
What is different now
The current transition shares features with past technological waves, but it also differs in important ways.
First, the scope of tasks affected is broader. Previous waves of automation disproportionately targeted routine, manual, or highly structured work. Today’s AI systems encroach on cognitive and creative tasks that were once considered safely human. This does not mean those jobs disappear, but it does mean their internal task mix changes.
Second, the cost curve looks different. Software and AI scale in ways that physical capital does not - granted, assuming the physical development of data centre infra is not constrained in the way some have argued. This creates the potential for faster diffusion, even if deployment remains uneven.
Third, the bottlenecks have shifted. Our conversations with CTOs across sectors highlight a consistent theme, that the technology often works, but the constraint is integration. Data quality, workflow redesign, accountability, and organisational trust matter more than raw model capability.
Fourth, the labour market is already more fluid. Job tenure has shortened, career paths are less linear, workers are expected to retrain, rebrand, and reorient repeatedly. This makes transitions both more common and more precarious.
There are early signs of strain, though they remain uneven and sector-specific. AI is still early and performs poorly at many tasks. But in areas where work is already digital and somewhat modular, the effects are beginning to show. The clearest early impacts are emerging in software engineering and in parts of the creative economy, particularly music, voice, and video production.
Software engineering
In software engineering, the shift so far is one of reconfiguration rather than collapse. Demand for experienced engineers remains strong, while junior roles appear increasingly squeezed as AI tools automate entry-level tasks such as debugging, documentation, testing, and boilerplate code.
A recent internal study by Anthropic, based on surveys of its own engineers, highlights the tension. Many reported productivity gains of around 50 percent, representing a two to three times increase compared to last year, with engineers able to fully delegate between 0 and 20 percent of their work to AI tools. Notably, 27 percent of AI-assisted work consisted of tasks that would not have been done otherwise, such as scaling internal tools or building quality-of-life improvements.
At the same time, some engineers expressed concern that their skills were becoming less sharp, and that the hands-on craft of coding was being eroded. Workplace dynamics are shifting too. As AI becomes the first stop for questions, engineers report turning to colleagues less often, raising questions about mentorship, collaboration, and how junior talent is developed.
These dynamics were echoed in a recent workshop we ran with CTOs, where several described deliberately using AI tools to compress or replace tasks traditionally assigned to junior engineers, improving short-term velocity but complicating long-term team development.
Creative industries
In the creative economy, impacts are arriving faster. AI-generated music, imagery, voice, and video are already widespread, from Spotify being increasingly populated with AI artists to advertising on the London Underground that increasingly relies on AI-generated visuals.
This is beginning to show up in data. A UK government report published in August found that after three consecutive years of growth, employment in the creative sector fell slightly. Survey evidence points in the same direction. A recent study led by Queen Mary University of London found that 73% of respondents believe generative AI is altering the quality of work in creative fields, while 68% reported reduced job security.
Hiring
Hiring is perhaps the key node in the critical path to addressing labour immobility. It is how people move from where they are to where opportunity exists. Yet it is also an industry that has changed remarkably little over the past decade.
The hiring sector remains fragmented across thousands of recruiters competing for completion fees, while job seekers are inundated with low-signal outreach and poorly defined roles. The result is a process that is noisy, inefficient, and poorly aligned with how work itself is changing.
This matters because hiring now sits at the intersection of several intensifying pressures. Labour shortages are widespread across Europe. The European Commission tracks labour shortages across most EU countries, particularly in construction, engineering, healthcare, ICT, and transportation and storage, including drivers and warehousing roles. These shortages are driven by a number of factors, that includes an ageing population, lack of mobility at borders and skills mismatches that prevent people from moving into available jobs. McKinsey estimates that in advanced economies, GDP in 2023 could have been 0.5 to 1.5 percent higher if employers had been able to fill open vacancies.
The challenge is not uniform. Blue collar work is more constrained by certifications and qualifications, creating an inherent inflexibility in this segment of the labour market. Think, qualifications for operating a forklift, electricians, plumbers, etc.
At the same time, the tools we rely on are failing. Job feeds on platforms like LinkedIn, Indeed, and TotalJobs are increasingly clogged with low quality slop. Employers struggle to articulate what they need. Job seekers struggle to assess relevance.
There are signs of progress. Welcome to the Jungle invested in job quality by reviewing employers and structuring job descriptions. Breakroom focuses on frontline workers, using employee-powered ratings to surface job quality rather than availability alone. Forage helps people explore career paths through simulations before committing to a role. More recently, companies like Jack and Jill combine AI-driven recruiting with career coaching, while Asap.work targets acute shortages in temporary construction work through a two-sided marketplace.
These approaches reflect a shared insight: Hiring is not just about filling roles, it is about translating skills, preferences and ambitions into genuine, desired opportunity. We think that hiring is one of the clearest fault lines in the workforce transition. Better hiring is not just an efficiency gain, but a core infrastructure for reskilling, mobility, and access to work.
This is also where responsibility matters. Under the EU AI Act, recruitment is classified as a high-risk use case for AI, requiring deliberate attention to bias, fairness, and accountability. We think AI is a huge enabler for delivering on the opportunity of enabling genuine labour mobility and long-term career satisfaction. But a lack of responsibility is the difference between delivering value to society and and not.
Below, we map the emerging landscape of companies that are re-building the hiring stack - this is a capture of some of the companies we think are genuinely doing something different in their segments:
Areas we’re excited about
We believe we’re entering a significant shift in jobs across the globe and across our economy. For now, the impacts on jobs are relatively small, but we anticipate this could ramp much quicker than many are expecting. And we see a massive need and opportunity for technology to unlock genuine labour mobility.
We are excited about teams that are building across three key domains.
Sector-specific, skills-based hiring platforms
Coaching and career transitions
Enabling a new wave of entrepreneurship
Sector-specific hiring platforms
This is an old category but we think is returning to the ring with vengeance. A new wave of hiring platforms is emerging that focuses tightly on specific sectors such as construction, logistics, healthcare, or renewable energy. The core thesis is that supply often exists, but demand and supply fail to meet due to fragmentation and reliance on generic hiring tools.
The value proposition is two-fold. Workers gain clearer visibility into relevant opportunities and can often access higher wages by cutting out intermediaries. Employers gain access to wider, more qualified talent pools that reflect sector-specific realities. Most platforms operate on transaction-based models tied to successful placements.
Within this, we are particularly interested in approaches that enable genuine skills-based hiring. While widely endorsed, skills-based hiring has failed to scale due to the cost and inconsistency of evaluating real capability. AI agents can shift hiring away from CVs and pedigree toward verified skills, but responsible data use and deliberate bias mitigation are essential.
Coaching and career transitions
Structural unemployment is increasingly a navigation failure. Many people lack visibility into what roles exist, what they could realistically be good at, and the trade-offs between different career paths. Today’s options are largely limited to expensive coaching or shallow online assessments.
AI enables continuous, personalised career guidance that helps individuals explore options, retrain, and transition repeatedly over time. As job tenure shortens, this becomes core infrastructure for labour mobility.
This category overlaps with re- and upskilling. Training providers often engage at the same moment as hiring platforms but focus first on building skills rather than placement. We distinguish between models that build on existing skills and those that support entry into new fields, and are particularly interested in approaches that combine training with clear incentives and credible routes into work.
Enabling a new wave of entrepreneurship
AI weakens the scale and bureaucracy advantages of large institutions, which we think will unlock a new wave of entrepreneurs. Could we eventually be living in a world of 5 billion entrepreneurs?
As the cost of building products and services falls, the potential for a new wave of local and specialised businesses increases. We are interested in platforms that act as the operating system for independent work. This includes handling finance, compliance, fundraising, HR, and risk, reducing the burden of operating outside traditional employment. In sectors where the employment contract erodes, entrepreneurship may become a meaningful source of economic security and labour mobility.
If you’re building in these spaces, whether currently just ideating, or have founded your company, we want to hear from you. Reach out on albert@revent.vc.
Notes
The phrase “great reallocation” is borrowed from recent economic research on trade and production dynamics. See, for example, Alfaro and Chor (2023) coining the “Great Reallocation” in global trade.
Photo by Mathias Reding on Unsplash