Key learnings from UN Week

AI and deep tech

1. AI's next bottleneck is physical.Cheaper inference now depends on hardware as much as on models. The firms pulling ahead design chips, servers and software together, and chips built only for transformer models trade general-purpose flexibility for much faster, more efficient inference.
2. Deep tech is where the next U.S. lead gets built.Robotics, quantum and space are the priority areas. Deployment was the theme: turning capability into practical value depends on infrastructure, integration and trust. New managers in space tech and quantum may have an edge over incumbents.
3. AI needs guardrails that leave room for small firms.Agents are moving from answering questions to running entire workflows, which raises hard questions for user experience and for industries like local news. Smaller firms still have an opening: in vertical AI, unique data and customer knowledge can beat scale.
4. Emerging managers give allocators returns and access.Established VC firms that back early managers show organizational maturity, and they see deal flow they otherwise wouldn't. The strongest new managers build from what they already have: an audience, relationships and an eye for talent.

Climate and emerging markets

5. Environmental damage is an economic problem for every investor.A Schroders synthesis of more than 40 reports puts longer-term nature dependency risk at about 50% of global GDP. Nature-related policy rose steeply from 2016 to 2024 (TNFD, the EU Deforestation Regulation, the EU Nature Restoration Law), and those costs are reaching company financials through regulation, insurance premiums and the cost of capital.
6. Impact measurement is becoming finance-grade.Schroders' NatCapEx model prices a company's impact on nature as a share of sales: a score of −2% means $2 of ecosystem services lost for every $100 of revenue, across 9,000+ listed companies. The Transition Pathway Initiative scores about 2,000 companies on management quality and benchmarks about 550 on carbon performance. No single metric captures corporate climate action, so measurement has to be tailored and data-heavy.
7. Cutting emissions is only part of the job.Protecting and restoring nature, removing carbon with technology and building resilience into supply chains matter as much.
   ~5 GtCO2e a year from tropical deforestation. Counted as a country, it would be the world's third-largest emitter, behind only China and the U.S.
8. The funding gap sits at the pilot and demonstration stage.Grants cover research and venture capital arrives after validation, which leaves first-of-a-kind projects stranded in between. One answer is a philanthropic revolving fund that pre-purchases carbon removal credits, resells them to corporate buyers on delivery and recycles the proceeds. The wider case is for "systemic capital," where debt, equity and philanthropy are combined on purpose.
9. Impact startups need help getting past Series A.Debt will be part of the answer. Better coverage and better data are needed to bring families and foundations in.
10. The Global South will draw more institutional and philanthropic capital.Local anchor capital is growing, and it pulls international allocators in behind it. Nigeria's sovereign wealth fund, NSIA, runs three ring-fenced funds and was accredited by the Green Climate Fund in March 2026 to scale blended climate finance. Local successes are reaching public markets too: the Dangote Refinery IPO, opened in September, is the largest in African history.