Yes — consolidation and capital are what the supply chain needs No — the bottlenecks are too deep for a holding company to fix Maybe — it depends on whether Onodrim can close deals at fair prices Too early — no acquisitions have been announced yet
Yes — regulatory approval and partner scale will make it happen No — the timeline is too aggressive given regulatory uncertainty Maybe — it depends on whether Part 108 passes on schedule Too early — Zipline hasn't proven it can operate at that scale yet
Yes — efficiency will determine how many tokens a data centre can process No — compute costs will dominate over power conversion Maybe — it depends on whether Vensum can prove its numbers at scale Too early — the company is still pre-revenue
Yes — human oversight will remain essential for complex risks No — AI will eventually underwrite with minimal supervision Maybe — it depends on whether regulators demand audit trails Too early — the market is still evolving
Yes — HF-free production will enable European sovereignty No — China's cost advantage is too entrenched Maybe — it depends on scaling to industrial production Too early — the company is still at pilot stage
Yes — validation will become a prerequisite for AI-driven security No — teams will deploy AI agents and fix data gaps later Maybe — it depends on whether regulators demand proof of detection Too early — the market is still forming
Yes — scale and speed will create defensible margins No — compute costs will keep margins thin across the industry Maybe — it depends on whether Fal can secure enough GPUs at the right price Too early — the inference market is still forming
Yes — agents will handle most routine B2B payment workflows No — humans will remain in the loop for financial transactions Maybe — it depends on whether enterprises trust agents with spending controls Too early — Pliant's agentic capability is still in beta
Yes — hardware-level suppression will prove more scalable No — software correction and scale will win Maybe — it depends on whether Quint’s architecture holds at hundreds of qubits Too early — the first chips haven’t been demonstrated yet
Yes — AI will eventually handle verification end-to-end No — human oversight will always be required for payments Maybe — it depends on whether AI can achieve zero false positives Too early — the company is still early-stage
Yes — pre-certified pods will make space research routine No — space will remain too complex for most labs Maybe — it depends on whether Oasis Beta delivers re-entry Too early — the first mission just launched in July
Yes — observation-based learning will replace manual configuration No — enterprises will prefer explicit, controlled automation Maybe — it depends on whether Marble can address privacy concerns Too early — the company just raised its first round
Yes — speed of execution will beat scale and incumbency No — primes and US rivals will ultimately win the contracts Maybe — Hypersonica will win a niche, not the main programme Too early — a prototype isn’t a weapon system
Yes — cost is the barrier, and Gallium is solving it No — traditional defense primes will maintain dominance Maybe — it depends on whether the six prospects convert to orders Too early — the radar works in tests, not in the field yet
Yes — new market structure favors firms built for it from day one No — scale and capital will keep incumbents dominant Maybe — LO:TECH could win specific niches like tokenised assets Too early — $12M buys offices, not market share
Yes — shared, multiplayer AI will become the dominant model No — single-user agents will remain the standard Maybe — it depends on whether Wabi can prove engagement and retention Too early — the company just launched publicly
Yes — AI will handle routine exceptions, humans will handle the rest No — procurement exceptions require human judgment Maybe — it depends on whether enterprises trust AI with purchasing decisions Too early — the technology is proven but adoption is still early
Yes — embedded connectivity will become standard in consumer apps No — telecom will remain a standalone service Maybe — it depends on whether businesses can make the economics work Too early — the market is still proving itself
Yes — simulation will prove more scalable than real-world data collection No — real-world data will remain essential for robust robot learning Maybe — it depends on whether simulation can model messy real-world edge cases Too early — the company has no commercial product yet
Yes — AI will handle most routine finance tasks within 5 years No — human judgment will remain essential for complex decisions Maybe — AI will augment, not replace, finance teams Too early — the technology is proven but adoption is still early
Yes — simulation will become standard before Phase II No — real trials will always be required for approval Maybe — it depends on whether the model's predictions hold up in human trials Too early — the first trial results aren't in yet
Yes — satellite data will become the primary compliance evidence No — supplier questionnaires will remain essential for verification Maybe — it depends on whether regulators accept satellite-based due diligence Too early — the EUDR enforcement hasn't started yet
Yes — integrated software and service will outcompete fragmented incumbents No — established players will adapt and maintain dominance Maybe — it depends on whether Cortas can scale service quality with software Too early — the company just signed its first customers
Yes — AI will handle most admin, freeing advisors for client work No — advisors will always need to oversee operational tasks Maybe — it depends on whether firms trust agents with custodian workflows Too early — the technology is proven but adoption is still early
Yes — virtual patients will become standard for early-stage testing No — real patients will always be required for approval Maybe — virtual patients will augment, not replace, real trials Too early — the technology is still proving itself
Yes — if it saves me money, I don't need to approve every switch No — I want to approve every financial decision Maybe — it depends on the category and the potential savings Too early — AI agents aren't trustworthy enough yet
Yes — AI will handle most routine procurement tasks No — human buyers will remain essential for quality and relationships Maybe — it depends on whether AI can handle complex negotiations Too early — the company just launched its agentic product
Yes — agent-native infrastructure will become a distinct layer No — hyperscalers will adapt and dominate agent workloads Maybe — it depends on whether 60ms provisioning becomes essential Too early — the agent cloud category is still forming
Onboarding — candidates drop out during slow, fragmented compliance Sourcing — not enough applicants to begin with Pay — workers leave for better wages elsewhere Retention — the first 90 days are the real challenge
Yes — AI-native firms will outcompete traditional accounting services No — trust and human judgement will keep traditional firms dominant Maybe — it depends on whether Gradient can scale quality with AI Too early — the company just raised its first institutional round
Yes — AI will become the standard for supply chain contractor compliance No — human judgement will remain essential for complex cases Maybe — it depends on whether HMRC accepts AI-generated evidence Too early — the company just raised its first institutional round
Yes — AI-native platforms will replace legacy property software No — traditional vendors will add AI features and maintain dominance Maybe — it depends on whether Vessla can prove ROI at scale Too early — the company just raised its first institutional round
Yes — Southern Europe will produce globally significant startups No — capital and talent will continue concentrating in Northern Europe Maybe — it depends on whether local funds like Enzo can scale Too early — the ecosystem is still maturing
Yes — precision additives will become standard for complex biologics No — standard media formulations will remain dominant for cost reasons Maybe — it depends on whether Silvia Bio can prove superiority at scale Too early — the company has validation but no named commercial contract yet
Yes — optical switching will become standard for GPU scale-up networks No — electrical switching will improve enough to remain dominant Maybe — it depends on whether optical manufacturing scales cost-effectively Too early — the technology is proven but deployment timelines are long
Yes — Ramp will become the default layer for AI spend management No — AI providers will build their own cost management tools Maybe — it depends on whether enterprises want a neutral layer Too early — the category is just forming
Yes — agent execution will become a core cloud primitive No — it will remain a niche feature within broader platforms Maybe — it depends on whether agents need persistent environments Too early — the category is still defining itself
Yes — AI-native firms will outcompete traditional services models No — relationships and reputation will keep incumbents dominant Maybe — it depends on whether Solvera can scale quality with AI Too early — the company just launched in July 2026
Yes — home batteries will become the backbone of residential grid flexibility No — utility-scale storage will dominate flexibility markets Maybe — it depends on smart meter rollout and device adoption Too early — the VPP business model is still proving itself
Yes — neutral governance will win because no provider can police rivals No — platform-bundled controls will be good enough for most enterprises Maybe — it depends on whether security teams demand cross-provider visibility Too early — the market is still defining the category
Yes — founder-led speed and customer focus can beat scale No — scale and capital will always win in telecom Maybe — it depends on whether AloFor can build repeat demand Too early — the company just launched in July 2026
Yes — creators will become the primary distribution channel for consumer software No — traditional product-led growth will remain dominant Maybe — it depends on whether creator audiences convert to paying users Too early — the model has traction but no scaled success yet
Yes — Heartbeat AI will become the universal energy OS for homes No — hardware will remain the core business Maybe — software will grow but hardware will dominate revenue for years Too early — the software pivot is still proving itself
Yes — two-sided agent networks will become the standard for B2B payments No — one-sided AP/AR platforms will adapt and maintain dominance Maybe — it depends on whether large enterprises adopt both sides Too early — the network is proving traction but scale is untested
Yes — AI decision tools will become essential for network management No — budget constraints and inertia will slow adoption Maybe — it depends on whether the M25 results can be replicated Too early — the market is still proving the value proposition
Yes — airships will become the standard for routine infrastructure inspection No — helicopters will remain dominant for complex or urgent missions Maybe — airships will capture routine inspections, helicopters will handle the rest Too early — manufacturing scale-up is untested
Yes — AI-native certification will become the standard for medtech No — traditional Notified Bodies will adapt and maintain dominance Maybe — it depends on whether regulators accept AI-assisted reviews broadly Too early — Scarlet’s model is proven but scaling is untested
Yes — AI testing will displace most manual QA outsourcing No — human testers will remain essential for complex scenarios Maybe — it depends on whether Klarent can scale verification economics Too early — enterprise adoption cycles are still early
Yes — the US capsule launch will happen on schedule 43% No — the timeline will slip again into 2027 35% Maybe — it depends on regulatory approvals 16% Too early — the company has no operational track record yet 4%
Yes — AI can transform service quality and resident experience 48% No — the biggest cost drivers are regulation and building safety 28% Maybe — it depends on whether Supa can scale beyond the first acquisitions 20% Too early — the company is just emerging from stealth 4%