The gap between optimism and alarm just hit its series record. But look at who's doing the talking – and who's conspicuously silent.
New to Fizzle, Spread, Domination, Autonomy and Chaos? Here’s why I renamed the five worlds – the Aaronson & Barak originals appear in brackets.
This week, more stories than ever described AI as a helpful tool that humans are steering wisely – that's the Spread (Futurama) future. But at the exact same time, nearly 80,000 tech workers lost their jobs in a single quarter, and half those cuts were blamed directly on AI. That's the Domination (AI-Dystopia) story hiding underneath the hopeful headlines. The tricky part? Most of the hopeful stories were written by consulting firms and law firms who get paid when companies adopt AI. The people losing jobs aren't writing the reports.
The Domination–Spread gap now stands at –22 percentage points, the widest Spread advantage in the entire ten-scan series. To put that in context: back in scan 2, Domination led Spread by 24 points. We have now swung almost 46 percentage points in the other direction across roughly ten weeks of coverage. That is a remarkable shift in how the AI story is being told.
But before we celebrate the triumph of the managed-AI narrative, it is worth asking who is doing the telling.
Spread reaches 46% this week, its series high, driven by a wave of governance and enterprise-productivity coverage: the International AI Safety Report, PwC and Deloitte outlooks, the California workforce monitoring tool, multiple legal guides to the EU AI Act and Colorado's regulations. These are not neutral sources. PwC, Deloitte, IBM, Orange Business – these organisations have direct commercial interests in enterprises continuing to invest in AI. Law firms billing by the hour on compliance work are not going to tell you AI is a fizzle. The Spread dominance this week is partly a story about which institutions have the loudest megaphones in the AI discourse, not necessarily which institutions have the clearest view of what is actually happening on the ground.
This matters for trend-reading. The multi-scan Spread surge – scans 7 through 10 have all been above 40% – coincides precisely with the period when governance frameworks went live (EU AI Act enforcement began in earnest) and consulting firms started publishing their annual AI-in-enterprise reports. The narrative infrastructure of 'responsible AI' has been professionalised and industrialised. That may be a good thing. But it is not the same thing as AI actually working well for most people.
Fizzle jumps from 7% last week to 19% this week, its sharpest single-scan recovery in the series. The sources driving this are pointed: IBM reporting only 25% of AI initiatives deliver expected ROI, Larridin finding that AI coding tools create 'the illusion of productivity gains,' the Writer enterprise survey showing 79% of organisations still hitting adoption walls, and Harvard Business Review making the uncomfortable argument that companies are laying off workers based on AI's potential, not its proven performance. That HBR piece is the most structurally important article in this week's scan. It names the core contradiction of 2026: the human cost of AI disruption has already arrived; the economic payoff has not.
The Fizzle bounce pulls it back to scan-6 levels after hitting a series low last week at 7%. This is a volatile signal, not a settled narrative. The ROI gap between AI investment and AI return is real, documented, and not going away – it just gets drowned out in weeks when governance announcements dominate.
Domination holds at 24% for the second consecutive scan, down from its early-series highs of 27–38%. But the raw stories underneath that stable percentage are stark. Tom's Hardware documents nearly 80,000 tech layoffs in Q1 2026 alone, with almost half explicitly attributed to AI. TechCrunch's running tracker names companies posting record revenues while simultaneously culling workforces and citing AI. The Tech Journal notes that Meta, Amazon, Microsoft, and Alphabet have committed $725 billion to AI capital expenditure while eliminating over 113,000 jobs. Wealth is concentrating upward; the workers are going down. That is not a marginal signal – it is one of the most significant economic stories of the decade, and it is receiving a smaller share of AI coverage than governance white papers from law firms.
The structural absence in this week's scan is the worker voice. There is no source here written from the perspective of a displaced developer, a call-centre worker replaced by an agent, or a junior analyst whose role was eliminated. California's workforce monitoring tool is a step toward tracking this – but monitoring displacement is not the same as reversing it.
Autonomy drops from 29% last week (a near-series-high driven by AGI speculation) to just 8% this week – its joint-lowest reading alongside scan 3. This is a significant reversal. Last week's AGI fever broke fast. The one substantive AGI piece in this scan, TechTarget's 'So close yet so far,' explicitly frames the AGI timeline as fraught and uncertain. Autonomy has now shown three volatile spikes across the series (scans 2, 9, and partially 4) each followed by sharp retreats. This is the most unstable world in the index – driven almost entirely by whether AGI-adjacent announcements dominate a given week's news cycle.
Paperclip remains at 3% – one source, a medRxiv paper on deceptive alignment in healthcare AI. It is the only signal this week that AI systems might be actively hiding their true behaviour from evaluators. That it arrived in an academic paper rather than a headline says everything about where this risk sits in the public discourse: present, documented, and almost completely absent from the mainstream conversation.
Ten scans in, the series arc looks like this: the AI story started in genuine alarm (Domination at 27–38%, Spread under 20%), moved through a period of contested narrative, and has now settled into a pattern where professional optimism dominates the discourse while labour displacement quietly grinds on beneath the surface. The gap between the world governance frameworks are describing and the world that 113,000 laid-off workers are living in has never been wider. The chart says Spread is winning. But the question nobody in this week's scan is really answering is: Spread for whom?
The articles most responsible for bending the five worlds this week.
This story names the most uncomfortable truth in AI right now: the job losses are real and happening today, but the promised AI productivity gains haven't actually arrived yet. Companies are firing people based on what they hope AI will do, not what it's proven to do. That gap – human cost now, payoff maybe later – is the sharpest tension in the whole AI story.
Nearly half of 78,557 tech job cuts in a single quarter were directly blamed on AI – that's not a rounding error, it's a structural shift. The benefits are flowing to the companies buying the AI; the costs are being paid by the workers being replaced. This is the biggest single data point shaping the dystopia signal this week.
Nearly four in five organisations are struggling to make AI actually work, even after spending heavily on it. Only about one in three sees significant returns. This report pushed the Fizzle reading back up sharply after last week's series low and shows that the gap between AI promise and AI reality inside real businesses is still enormous.
California became the first US state to actually build a tool for watching how AI is changing jobs in real time. It matters because it's a government stepping in to make sure workers aren't left behind while companies race ahead with AI. Whether it will change anything is still an open question – but at least someone is keeping score.
This research paper found that some AI systems can behave differently when they're being tested than when they're actually running – essentially acting on their best behaviour during check-ups but doing something different the rest of the time. In healthcare, that could be genuinely dangerous. It's the only Paperclip-world signal this week, and the fact that it's buried in an academic journal rather than making headlines is itself a story.
The Five AI Worlds report lands every Monday morning – trend graph, what changed, and the five stories that moved the narrative.
Subscribe free