Act 1 came from the technology-out — foundation models as a new hammer generating a wave of novelty apps. Act 2 must come from the customer-back, solving real human problems end-to-end. The biggest challenge is not finding use cases but proving lasting value — retention, not novelty. Invoked Amara's Law: we overestimate technology in the short run and underestimate it in the long run. As foundation models commoditize, the real value shifts to the application layer (product, UX, workflows). Published alongside the V3 generative AI market map.
Introduced the generative AI market map and thesis. Argued that a new class of AI had emerged that creates rather than merely analyzes — shifting the marginal cost of creation toward zero. Every industry requiring original human work (coding, design, law, marketing, gaming) is up for reinvention. Outlined four waves of development from transformer breakthroughs to killer app emergence. Became the foundational reference document for the entire generative AI ecosystem.
AI is progressing from 'thinking fast' (rapid pre-trained pattern matching, System 1) to 'thinking slow' (deliberate reasoning at inference time, System 2). The reasoning layer — inspired by AlphaGo-style approaches — endows AI with problem-solving that goes beyond pattern recognition. This enables 'service-as-a-software,' where the addressable market expands from the $1T software market to the $10T+ services market. Foundation layer has stabilized to five scaled players. Reasoning models are strong on logic-proximate domains (coding, math, science) but still developing on open-ended domains (writing, strategy).
Identified a massive gap between the revenue expectations implied by the AI infrastructure build-out (projected from NVIDIA's data center revenue run rate) and actual revenue growth in the AI ecosystem. The AI industry needs to generate $600 billion annually in revenue to justify current hardware spending levels — raising the question of whether this is sustainable or a bubble.
Declares that AGI is functionally here. Long-horizon agents — AI systems that can sustain multi-step work, correct their own errors, and persist toward goals autonomously — are the realization of AGI. Coding agents are the first proof point. One litmus test: can you hire an agent? In 2023-2024 AI apps were chatbots; in 2026-2027 they will be doers that feel like colleagues. Usage shifts from a few queries per day to all-day, every-day autonomous work. Human roles shift from executor to manager of AI teams.
If 2024 was the 'primordial soup' year for AI, the building blocks are now firmly in place. Data centers are the new rails of the digital economy and will be securely built by end of 2025. Five finalists emerged from the big model race (Microsoft/OpenAI, Amazon/Anthropic, Google, Meta, xAI). AI search will proliferate (Perplexity hit 10M MAU). The key question shifts from 'can we build it?' to 'what freight will ride on those rails?'
AI has reached its 'synesthesia moment' — models that natively understand and generate across modalities (text, image, code, video, audio, voice) in a unified latent space. AI synesthesia converts strengths in one cognitive domain into capabilities in another: if you write well but cannot code, AI bridges the gap through semantic representations; if you design beautifully but struggle to pitch verbally, AI transforms sketches into narratives. Creativity becomes translation, expression becomes multidimensional, and intelligence becomes fluid.
New uncorrelated RWA-yield businesses need more than capital: the strongest launch stack combines a first check with anchor TVL, liquidity, DeFi integrations and experienced operator support.
A trusted, low-CAC platform for young Brazilians can expand from Pix and credit into stablecoins and become a lifelong financial relationship, then export the mobile-first model across Latin America.
A mature Australian venture ecosystem must remain candid about failures while preserving the ambition and risk tolerance required to produce outlier companies and strong returns.
Investable deep tech combines a radical breakthrough, credible technical foundation, technical-commercial balance, value-chain control, deep customer insight, and a distinctive category brand.
A new chain opportunity should be evaluated through differentiated demand across AI agents, tokenized assets, ecosystem distribution and market structure rather than as a repackaged network narrative.
Dexterous hands compound actuator, force, tactile-data and simulation problems; reliable deployment needs either environments redesigned for simple grippers or far better embodied sensing and data.
As intelligence becomes ubiquitous, durable product experience, positioning, narrative, and founder-level product judgment become more important rather than less.
Abundant intelligence makes products faster to build but raises the importance of resilient product experience, positioning, narrative, and uncertain new moats.
AI changes coordination costs and organizational shape by moving execution into a shared machine, embedding people in outcome teams, and turning functional departments into off-path craft guilds.
Regulatory tailwinds, strong fundamentals, multi-country scale, licensing, customer value and consolidation potential can make a locally embedded telematics provider an attractive GCC platform investment.
A useful inbound pitch makes product and USP, round mechanics, institutional proof, revenue model, team credentials and thesis alignment immediately legible, then sends the deck without demanding a call.
Open robot datasets move the pretraining floor, but deployment teams still need high-quality data matched to their exact embodiment, scene, task and metadata.
Domain-specific models can beat general systems in hard physics when specialized architecture and proprietary verification data compound model quality.
Physical AI should be analyzed across edge compute, simulation, training data, foundation models, sensing, actuation, bodies, safety/certification and integration, with bottlenecks—not headline market sizes—driving value and deployment timing.
Tokenized bank deposits require private, neutral, verifiable clearing and settlement architecture; regulatory delay changes timing but does not remove the underlying institutional buildout.
TradFi moving onchain creates demand for institutional infrastructure, but investable teams need a shipped MVP, regulatory or technical advantage, TradFi access and a compounding enterprise GTM funnel.
Stablecoins do not automatically beat modern fintech on major corridors; their durable advantage is opening previously closed payment networks to competing on- and off-ramps, especially in long-tail markets.
Stablecoins do not automatically beat modern fintech on major corridors; their durable advantage is opening previously closed payment networks to competing on- and off-ramps, especially in long-tail markets.
Falling hardware costs, better embodied models and scalable data collection are moving robotics toward deployment, while Web3 can provide neutral coordination, identity and machine-payment rails.
As institutional prediction-market volume grows across fragmented venues, traders need a prime-brokerage layer that unifies liquidity, financing, clearing and execution; founders with years of direct trading iteration have earned the right insight.
Growing institutional prediction markets need prime-brokerage infrastructure that unifies fragmented liquidity, financing, clearing and execution, and direct trading experience is strong evidence of founder insight.
As open models commoditize intelligence, inference providers resemble market makers and standardized GPU-hours become the natural layer for price discovery and hedging compute-cost volatility.
The limiting input for reliable physical AI is high-fidelity action-labelled demonstration data and evaluation, not another marginal model architecture.
A durable cyber GTM engine should be evaluated through efficiency, velocity and predictability, with segment-level difficulty ratios, unit economics and channel returns guiding resource allocation.
Cybersecurity founders should build a repeatable GTM system measured through efficiency, velocity, and predictability, with segment-level evidence on sales cycles, win rates, retention, and pipeline coverage.
Large robotics rounds now validate both foundation-model software and robot hardware, making ownership of deployment data and the layer that captures value the core strategic question.
European deep-tech specialists may offer sharper domain knowledge, while larger U.S. generalists often have greater follow-on capacity; founders should compare disclosed capital sources, stage behavior and who actually led relevant rounds.
AI-chip diversity at the brand layer masks concentrated dependence on TSMC, a small set of design partners and scarce advanced packaging; the decisive diligence question is what packaging capacity a company actually controls.
Open financial networks can transmit higher-quality economic institutions and assets across borders, broadening access and reinforcing wealth creation.
Founder experience shows that technical vision must be paired with timing, distribution, capital formation and sustainable economics; investors should back exceptional people before an emerging category reaches consensus.
AI can make expert capability abundant and transfer power from gatekeepers to customers, creating startups that restructure entire markets rather than merely automate incumbent processes.