Today on The Send: a confirmed BLM director who's spent 30 years trying to sell off public lands, the Forest Service handing 193M acres to states, Hilton calling a C-shaped consumer recovery while sentiment hits a 48-year low, and YC's Summer RFS pivoting hard into hard tech.
The Senate confirmed Steve Pearce as BLM Director on April 29 via en bloc vote, burying his nomination among dozens of others. Pearce's 30-year congressional record includes opposing public lands protections, advocating land disposals, attempting to shrink national monuments, and supporting oil and gas expansion β positions he declined to renounce at confirmation. He now oversees roughly one in every ten acres in the U.S. The same week, the White House completed review of the Biden-era BLM conservation rule (moving toward revocation) and approved a categorical NEPA exclusion accelerating geothermal exploration permits.
Why it matters
Pair Pearce's confirmation with the Forest Service decentralization story below, the still-unfilled NPS directorship, and the conservation-rule rollback, and the federal lands operating model is being structurally rewritten in a single quarter. For anyone building in guided travel or outfitting, the operational implications stack: extractive uses prioritized over recreation in BLM planning, conservation leasing as a funding mechanism likely gone, and state/tribal stewardship picking up the slack unevenly. Permit predictability β the actual variable that shapes guide-business unit economics β gets worse before it gets better.
The operating-model layer of the Forest Service restructuring now has a name and a mechanism: Chief Tom Schultz (former timber executive) is pushing 20-year shared stewardship agreements handing significant management responsibility for portions of the 193M-acre national forest system to states and tribes. This lands on top of the previously reported Salt Lake City HQ relocation, elimination of all 10 regional offices, and the move to a 15-state-director model by mid-2027 β a management structure not used since 1907.
Why it matters
Prior coverage established the structural shape (HQ relocation, regional-to-state-director model, 22% trail maintenance drop, 50 research station closures). Today adds the explicit devolution mechanism: 20-year intergovernmental agreements that will make permit issuance, trail standards, and access rules diverge by state. That patchwork is the variable that most directly determines guide and outfitter unit economics β operators with strong state and tribal government relationships gain a durable advantage; national platforms relying on uniform federal frameworks face fragmentation risk. Idaho, Montana, and Alaska are the likely first-mover templates to watch.
Two more sites join the managed-access stack. Arizona's Slide Rock State Park rolls out a batch-fill entry system May 1, eliminating first-come queuing and admitting vehicles only as its 156 parking spots turn over. New York's DEC released a report documenting record post-2020 Adirondack visitation and trail damage, proposing a hiking reservation system with capacity caps and no-parking zones β a formalization of the previously reported ~400/day Adirondack Loj and ~240/day Cascade Mountain caps β now facing local pushback over advance-booking friction. Oregon separately approved drone bans across most state parks and all 363 miles of coast effective May 1.
Why it matters
The pattern of parking-as-permit, advance reservations, and activity-specific bans is now executing at the state level in the same week the federal capacity engine (NPS down 25% staff, no confirmed director) breaks down. The Adirondacks pushback is a new data point not in prior coverage: spontaneous-access expectations remain a live political constraint that will slow implementation timelines, which is relevant for any booking or operator tooling built around capacity-aware dispatch.
Oman's Ministry of Heritage and Tourism issued the Executive Regulation of the Tourism Law on April 28, requiring all tourism businesses β including tour guides, adventure tourism operators, travel agencies, and hotels β to obtain government licenses within six months. The framework replaces decree 124/2021 and is paired with a new centralized AI-powered tourism data hub. North Carolina separately filed SB 837 mandating $1β2M liability insurance for all motocross facilities (effective July 2027), signaling parallel formalization of adventure-sport liability frameworks at the U.S. state level.
Why it matters
Read together with Tasmania's $100 enthusiast licence, Vietnam's national adventure standard, and the NC motocross bill, this is the consistent direction of travel: adventure tourism is being formalized into licensed, insured, professional categories everywhere. That raises floor costs for informal operators (good for digitized booking platforms with compliance tooling baked in) and creates real competitive moats for platforms that solve KYC/insurance/permit overhead β the BirdDog thesis applied across geographies.
Booking Holdings reported Q1 2026 gross bookings up 15% to $53.8B, with AI assistants (Priceline's Penny, Booking.com's natural-language search, OpenTable's AI Concierge, Agoda automation) driving measurable conversion uplift and double-digit cost reductions per booking at Agoda. CEO Glenn Fogel framed AI as a competitive moat against general-purpose AI assistants. KAYAK launched its own Ask AI conversational layer the same week. Yatra's DIYA assistant added multi-city itinerary planning.
Why it matters
This is the OTA establishment's response to the Anthropic/OpenAI travel-integration thread covered earlier in the week β and it's working. The $53.8B platform is using its data and supply moat to make conversational booking a defensive feature, not a disruption vector. For outdoor and adventure operators, the implication is that the booking layer is consolidating around AI-native UX faster than expected, raising the bar for any operator-direct or vertical-specific marketplace play. Conversational search is now table-stakes; the differentiation has to live elsewhere β supply curation, local relationships, or last-mile workflow.
Netflix co-founder Reed Hastings launched phase two of his Powder Haven private ski community at Powder Mountain, Utah (8,000+ skiable acres). The new 34-lot Prado neighborhood is priced at $4M+ per lot and includes two new chairlifts. The structural play: revenue from the private community subsidizes operations of the public resort underneath.
Why it matters
Pair this with Michael B. Schwab's wave-pool real-estate empire (Cabo Real Surf Club, $200M+ presales) and Saudi Arabia's ADRENA build, and the pattern is unmistakable: the highest-return model in adventure-sports infrastructure right now is luxury real estate stapled to skill-development venues, with the public-access experience monetized as an amenity for landowners. It works because residential lot economics ($1.4Mβ$4M per unit) underwrite the capex that traditional resort P&Ls can't justify. For founders, the question is whether software/platform plays can attach to this model or whether they get squeezed out by integrated operators owning land, infrastructure, and customer.
China's commercial climbing-gym count grew 31% YoY to 636 facilities by end of 2023, with climber-entrepreneurs opening chains and high-profile investors like Vanke founder Wang Shi backing startups. Unit economics show 2β5 year payback cycles. Reporting frames the sport as transitioning from university niche to youth-driven fitness category, with parallel build-out of coaching standardization, training systems, and equipment brands.
Why it matters
Counter-cyclical to the UK climbing-gym bifurcation story (11β15% revenue drop, mid-tier squeezed) covered earlier this week. Two markets in different stages of the same maturity curve: UK consolidating, China still in expansion. The cross-cut is useful for anyone modeling a global climbing-platform thesis β coaching/junior pipelines and data analytics are the survival levers in mature markets, while pure facility counts still drive growth in emerging ones. Olympic visibility and social-media-driven discovery are doing the demand work in both.
Y Combinator released its Summer 2026 Request for Startups across 15 categories, with a clear shift toward hard tech (agriculture robotics, defense, semiconductors, space) while preserving seven software-first categories. VC Cafe's synthesis pairs the YC RFS with a16z Speedrun's agent-network thesis and ARK's macro framework, identifying four converging patterns: agents that become networks, company operating systems, software built for agents (not humans), and AI-native services selling outcomes. Q1 2026 data backs the pivot β global VC hit a record $330.9B but with the lowest deal count in five years, AI capturing 58β80% depending on the cut, and seed-to-Series-A conversion at just 1.2% for 2023 vintages.
Why it matters
Read together with Elad Gil's '12β18 month exit window' framing covered earlier in the week, this is the clearest single signal of where venture capital expects the next set of outsized outcomes to come from: physical, regulated, and agent-native categories. For a second-time founder evaluating outdoor/travel adjacencies, the takeaway is encouraging β capital is actively repositioning toward businesses that touch hardware, regulation, or vertical workflow ownership, which is exactly the shape of an outfitter/guide platform with real supply-side relationships. The 'missing middle' problem (Australia, India, Africa data this week) means execution proof from day one matters more than narrative.
Actively AI, founded by 26-year-old Mihir Garimella, raised a $45M Series B co-led by TCV and First Harmonic at a $250M valuation. The product builds custom AI agents for sales teams that autonomously research accounts, draft outreach, and suggest next steps. Early customers Ramp and Verkada report meaningful productivity and revenue gains. The win mechanic: integrating into existing tools (email, Slack, Salesforce) rather than replacing them. Backstops the same week's Avoca news ($125M Series B at $1B for HVAC/plumbing voice agents) and Parallel Web Systems ($100M Series B at $2B for agent-native web infrastructure).
Why it matters
This is the NEA 'last-mile' thesis covered earlier in the week β defensibility lives in workflow ownership, not model differentiation β playing out at speed across SaaS, field services, and infrastructure. The repeating pattern: AI agents that wrap and replace specific human workflows in established categories reach unicorn valuations in 12β24 months. For a founder scoping outdoor/travel verticals, the operator-side equivalents (booking ops, dispatch, post-trip review automation, guide-management ops) are the cleanest analogs. The fact that incumbents like Salesforce face real displacement risk from purpose-built agents is the most useful signal here.
OpenAI open-sourced Symphony, an agent orchestration spec that turns Linear issue trackers into control planes for Codex coding agents. Every open ticket gets a dedicated agent running until completion; teams report a 500% increase in landed pull requests in three weeks. Implementation is a single SPEC.md file, allowing non-engineers to file requests and receive working implementations. JetBrains separately announced its 2026 Agent Client Protocol (ACP) so external coding agents (Cursor, others) can integrate directly into JetBrains IDEs β making IDE environments agent-agnostic.
Why it matters
The Symphony pattern is the most concrete data point yet on what 'AI as substrate' means at the team level β not 'developers write code faster,' but 'product managers file Linear tickets and ship working PRs.' For a small founding team, this collapses the boundary between PM and engineering in a way that meaningfully changes hiring math. Combined with JetBrains' move to agent-agnostic IDEs and IBM Bob's full-SDLC orchestration play, the developer-tooling stack is rapidly becoming a multi-agent orchestration layer rather than a code-writing assistant. Worth setting up a Symphony test before the pattern becomes table-stakes.
Hilton CEO Christopher Nassetta argued improving mid- and lower-chain hotel performance signals the K-shaped economy is converging into a 'C-shape,' citing falling inflation, expected rate cuts, and AI productivity gains reaching lower-income consumers. This lands in direct tension with the broader macro context: University of Michigan consumer sentiment hit 49.8 in April β the lowest since 1978, below 2008 and pandemic lows, and below the prior reported 47.6 floor β while consumers continue spending by drawing down savings and floating expenses on credit. Visa beat Q1 estimates on resilient card volumes; UBS and Piper Sandler describe a 'Wile E. Coyote effect' where war/oil-price impacts haven't yet hit. European summer travel-planning intent sits at 82%; Indian youth report 55% adventure-first preferences.
Why it matters
Prior coverage flagged extreme partisan divergence in sentiment (Republicans 87.1, Democrats 31.8) and the $181.6T household wealth effect as a spending prop. The new element today is Nassetta's C-shape thesis β a concrete counterargument from an operator with live pricing and occupancy data across income tiers. The Wile E. Coyote framing is also new: a named mechanism for why the gap between sentiment and actual spending closes suddenly rather than gradually. For founders pricing 2027 demand, the operating answer is still to model both scenarios β soft landing (good for mid-market guided travel) and Q3βQ4 credit-drawdown exhaustion β but the Hilton data point is the first institutional voice arguing the bottom cohort is recovering, not just the top.
The CFPB finalized amendments to Regulation B, effective July 21, 2026. The changes eliminate the 'effects test' for disparate-impact claims under the Equal Credit Opportunity Act, narrow the 'discouragement' prohibition to oral and written statements only, and impose new restrictions on special-purpose credit programs. The move aligns with the administration's broader effort to reduce fair-lending supervision and enforcement.
Why it matters
For anyone with active fintech exposure, this is the single most consequential regulatory change of the cycle on the underwriting side β disparate-impact has been the primary federal vector for fair-lending challenges to algorithmic and AI-driven credit models. Its removal at the federal level reshapes risk calculus for AI underwriting (relevant to today's CFTC AI-review story and last week's Affirm agentic-credit launch). Two important caveats worth tracking: state fair-lending laws and the Fair Housing Act still impose stricter standards, and the political durability of this rollback past 2028 is far from certain. Anyone running SPCPs targeting underserved populations should re-paper documentation now.
The public-lands attrition story moves from threat to operating reality Pearce confirmed at BLM, Forest Service handing 193M acres to states/tribes via 20-year stewardship deals, BLM conservation rule under revocation, geothermal NEPA exclusions approved, and NPS still without a director heading into peak season. The architecture of federal stewardship is being restructured, not just defunded.
Capital is bifurcating: mega-rounds at the top, seed starvation at the bottom Q1 global VC at a record $330.9B but with the lowest deal count in five years; AI takes 58β80% depending on the cut; India fintech rounds halve while late-stage funding climbs 126%; Africa equity drops 27% as debt 6x's. The 'missing middle' is now visible across every region.
The K-shape vs. C-shape consumer debate is getting concrete Hilton's CEO calls a C-shaped convergence as mid- and lower-chain hotel performance improves; Michigan sentiment hits a 48-year low while consumers float spending on credit; Visa beats on resilient volumes. The signal mix is incoherent enough that founders pricing 2027 should plan for both scenarios.
AI moves from tool to substrate β and YC officially pivots away from pure software YC Summer 2026 RFS goes hard-tech (agriculture robotics, defense, semiconductors); Avoca hits $1B on field-services agents; Actively AI raises at $250M challenging Salesforce; IBM ships full-SDLC orchestration. The repeating thesis: AI is the substrate, defensibility lives in vertical workflow ownership and the 'last mile.'
Adventure and experiential travel keep printing demand even as macro wobbles Europe summer planning at 82%, India youth at 55% adventure-first, ATTA's $185B North America adventure market, gear market projected to $79.5B by 2031. Booking platforms are wiring AI in (Booking Holdings, KAYAK Ask AI, Yatra DIYA) while regulatory formalization (Oman licensing, NC motocross insurance, Oregon drone bans) raises operator costs.
What to Expect
2026-04-29—Fed holds rates at Powell's likely final FOMC meeting; Warsh transition begins. SoFi Q1 earnings.
2026-04-30—Connecticut DEEP's inaugural Outdoor Recreation Day at the State Capitol β 40+ exhibitors mapping the state-level outdoor economy.
2026-05-01—Bonsoy Gold Coast Pro (Snapper Rocks) opens β third stop of the WSL CT under the new accumulated-points format. Slide Rock State Park's batch-fill entry system goes live.
2026-05-08 to 05-10—IFSC speed climbing competition in Wujiang follows Yicheng Zhao's new 4.58s world record.
2026-07-21—CFPB's Regulation B amendments take effect β disparate-impact test eliminated under ECOA.
How We Built This Briefing
Every story, researched.
Every story verified across multiple sources before publication.
🔍
Scanned
Across multiple search engines and news databases
832
📖
Read in full
Every article opened, read, and evaluated
174
⭐
Published today
Ranked by importance and verified across sources
12
β The Send
π Listen as a podcast
Subscribe in your favorite podcast app to get each new briefing delivered automatically as audio.
Apple Podcasts
Library tab β β’β’β’ menu β Follow a Show by URL β paste