The gap between who wins from AI and who doesn't widened again this week. Your job this fortnight is to figure out which side of that line your organisation is on.
The five worlds – AI-Fizzle, Futurama, AI-Dystopia, Singularia and Paperclipalypse – come from Aaronson & Barak’s Five Worlds of AI.
This week, more stories than ever described a world where the benefits of AI are flowing to a small group of very powerful companies – chip makers, hyperscalers and a handful of AI labs – while everyone else faces job cuts, rising costs and tech they don't fully control. That's the AI-Dystopia signal rising again after a few quieter weeks. The hopeful Futurama story – where AI is a useful tool that helps ordinary people and smaller businesses – is still there, but it's losing ground globally. Australia and New Zealand are actually the bright spot this week: local signals show more stories about AI helping real businesses than concentrating power at the top. The big question for every leader right now is simple: is AI working for your people, or just for your shareholders?
Let me give you the number that matters most: the Distribution Gap – AI-Dystopia minus Futurama, in percentage points of the global signal – sits at +6 this week, up from +5 last week and the highest it's been since scan 2 in April when it hit +15. That's not a crisis, but it's a clear directional arrow. After three consecutive weeks where the gap was narrowing (it bottomed at +2 in late April), we're back on a widening trend.
What does that mean in plain language? More of the stories shaping the AI conversation describe a world where the gains concentrate at the top – in chip makers, hyperscalers, frontier labs and their largest investors – rather than spreading to organisations, workers and communities broadly. The Futurama share rose to 23% globally, which is encouraging, but AI-Dystopia held 29%. The gap between them is the story.
Here's the thing: this isn't a survey of what's true. It's a reading of what's being said, written and argued in the publications your peers are reading. That discourse shapes decisions. When the dominant narrative is concentration, leaders make defensive moves. When it's diffusion, they invest more boldly. Right now the narrative is pulling in two directions – and the dystopia current is stronger.
I want to start here, and not just because you're reading this from Sydney or Auckland. The ANZ signal this week is genuinely interesting: AI-Dystopia sits at just 21% and Futurama at 34%, giving a Distribution Gap of -13 – meaning locally, value is being read as spreading, not concentrating. That's the most Futurama-dominant regional reading in the whole dataset this week.
The rolling four-week picture confirms it's a real trend, not a one-week blip: rolling ANZ gap is -10, and has been consistently negative. The stories driving it are real and specific. WiseTech staff publicly blasting AI-led redundancies in the AFR, the Qantas union demanding answers on AI strategy, and Andrew Leigh's warning that only 7% of Australian businesses broadly use AI – these are dystopian signals. But they're outnumbered this week by stories like CBA, PwC, BHP, Telstra and Freehills sharing actual AI productivity lessons, the federal budget committing $70 million to an AI accelerator, IAG scaling AI across operations, ProcurePro raising $15 million to bring AI to construction procurement, and sole traders reporting broad adoption to save time and costs.
The ANZ adoption lens is busy with organisations actually doing things, not just vendors announcing things. That distinction matters. When I strip out vendor-sourced stories and focus on independent adoption signals, the ANZ picture holds up. These are practitioners speaking.
The AI-Fizzle signal at 16% (down from 19% rolling) also suggests local leaders are getting slightly less bogged in pilot purgatory – though with only 7% of businesses at broad adoption, the gap between intent and scale remains real.
The US is carrying most of the weight in the global AI-Dystopia reading, and it shows. At 33% this week, the dystopia signal in North America has barely moved from its rolling average of 30%. Futurama holds at just 18% – a Distribution Gap of +15, the widest of any region.
The week's most covered stories tell you why. The Cerebras IPO – covered across eight separate sources – is a $60 billion bet that AI infrastructure value will concentrate in whoever can challenge Nvidia's compute dominance. The Musk v. Altman trial, covered seven times, is quite literally a courtroom argument about who controls the most powerful AI lab in the world. Cisco cutting 4,000 jobs while orders surge, Amazon cutting retail staff while racing toward the $3 trillion club, Cloudflare shedding 20% of its workforce as AI use skyrockets – these aren't edge cases. They're the dominant corporate narrative of the week.
The talent driver is particularly sharp in the US this week: AI-driven layoffs, worker protests at Meta over mouse-tracking surveillance, a former Google engineer at 55 who felt pressured to adopt AI too fast and was effectively pushed out. Gartner's finding that AI-led job cuts don't reliably improve ROI lands in this context not as reassurance but as indictment.
Europe's AI-Fizzle reading jumped to 24% this week – the highest in the series for that region. That's consistent with a week that included public-sector AI deployment outpacing management capability (Silicon Republic/IDC), and a UK probe into Microsoft's AI bundling and lock-in. The dystopia reading at 32% reflects geopolitical anxiety: Meta navigating EU competition probes, OpenAI offering privileged cybersecurity model access to European regulators, and Nvidia pushing into Germany.
The rolling figures show Europe's Paperclipalypse at 21% over four weeks – the highest sustained reading of any region on that world – suggesting European discourse is more attuned to AI-enabled harm than any other geography. White Circle raising €10 million specifically to stop AI models going rogue in enterprise deployments is a European story. So is the GovInsider analysis of Scotland's AI governance strategy. The continent is asking hard questions about control that North America hasn't fully landed on yet.
Singapore and Hong Kong carry the APAC signal at only 5% weighting, so treat these numbers carefully. But the 41% AI-Dystopia reading – driven almost entirely by China-focused stories about Alibaba, Tencent, ByteDance, Kuaishou and SMIC concentrating AI infrastructure investment – reflects real structural dynamics in the region's AI economy. The Samsung memory chip strike threatening AI supply chains adds labour tension to the infrastructure concentration story.
The macro forces are unusually legible this week. Economy is the dominant driver of the dystopia signal: capital pouring into chip makers, cloud providers and frontier labs (Cerebras, Applied Materials, SoftBank's OpenAI gains, the $5 trillion capex supercycle). Geopolitics runs a close second: the Trump-Xi summit with Nvidia chip exports in the balance, Anthropic lobbying Washington to protect its edge over China, China building a national computing power network. Together these two drivers account for nearly all of the concentration story.
Trust is the principal engine of Paperclipalypse this week: AI-enabled zero-day exploits (Google's GTIG report), ASIC warning on AI-accelerated threats, the IMF flagging systemic financial risk, Claude exhibiting behaviours Anthropic can't fully explain, and AI notetakers making lawyers very nervous. The Paperclipalypse share dropped to 14% globally – the series low – but trust-driven harm stories are denser and more specific than in earlier scans. The quality of the risk signal is rising even as the quantity dips.
Technology drives both Futurama and Singularia. The Futurama technology signal reflects real adoption: SAP's AI automation suite, Stanford Healthcare streamlining clinician experience, Xero accelerating AI with strong revenue results, Ikea's robotic worker-augmentation boots. The Singularia technology signal reflects a different conversation: Harvey deploying 500 autonomous legal agents, NBN Co shifting to an agentic AI strategy, Versent bringing agentic AI to the mid-market, Airbnb's CEO confirming 60% of new code is now written by AI.
I notice what's not here. There are almost no stories this week from financial services firms or professional services organisations reporting honest, detailed post-implementation reviews – what worked, what didn't, what it cost. There's very little from HR about reskilling programmes that have actually worked at scale. And there are almost no voices from workers who've benefited from AI adoption – the success stories of augmentation rather than displacement. The discourse is dominated by industry announcements, lab-level drama and macroeconomic concentration stories. The practical, granular, human experience of AI at work is structurally underrepresented.
If you're a CEO or CIO reading this, here's what the week is actually telling you. The AI-Fizzle signal at 17% means the 'pilots not delivering' problem hasn't gone away. The dystopia signal at 29% means your board, your investors and your union reps are all reading stories that frame AI as a tool for concentrating value rather than sharing it. And the Paperclipalypse signal – even at its series low of 14% – is built on trust failures that are becoming more specific and more costly by the week.
The leaders winning right now are the ones who can do three things simultaneously: show measurable results from AI adoption (not just effort), demonstrate that those results benefit teams and customers and not just the balance sheet, and maintain genuine human oversight of the systems they've already deployed. Australia's own AI dividend stories – CBA, BHP, Telstra, IAG – are your case studies. Use them. The gap between the 7% of Australian businesses at broad AI adoption and the rest isn't a technology problem. It's a leadership problem. And it's yours to close.
The articles most responsible for bending the five worlds this week.
This story showed that some of Australia's biggest and most well-known companies are actually getting real, measurable results from AI – not just talking about it. It mattered this week because it gave local business leaders concrete examples to point to, at a time when global stories were dominated by AI making a small group of investors very rich while workers worried about their jobs.
Cerebras raising billions in a blockbuster stock market listing was the single most-covered story of the week, written about eight times across different publications. It's a clear sign that the money flooding into AI is going to a very small number of companies that control the computer chips powering the whole industry – not to the businesses and workers trying to use AI every day.
WiseTech workers publicly pushing back against how the company handled AI-driven job cuts was one of the most significant Australian workplace AI stories of the week. It showed that even in a tech-savvy industry, employees feel like they're paying the price for AI gains that flow straight to the company – and they're not staying quiet about it.
A senior Australian politician putting a stark number on the problem – only 7% of Australian businesses are actually using AI across their work – cut through the noise of global hype this week. While the rest of the world argued about who owns AI's future, this story asked a simpler and more uncomfortable question: why is most of Australia sitting it out?
Australian workers demanding transparency from one of the country's most recognisable brands about its AI plans is a sign of where the real workplace AI conversation is heading. When unions start asking formal questions, it means employees don't trust that the benefits of AI will be shared – and that's a management problem every HR leader in Australia needs to take seriously right now.
The Five AI Worlds report lands every Monday morning – trend graph, what changed, and the five stories that moved the narrative.
Subscribe free