No PriorsNo Priors Ep. 119 | With Applied Intuition's Qasar Younis and Peter Ludwig
CHAPTERS
- 0:00 – 1:58
Founding mindset: staying quiet, capital-efficient, and focused on execution
Elad and Qasar open with light banter that quickly turns into why Applied Intuition stayed unusually “quiet” despite rapid growth. Qasar explains the strategic value of keeping a small public identity early on to avoid locking into a prematurely-defined narrative and to preserve operating flexibility.
- •Applied Intuition’s rapid scale: profitable, large team, significant revenue, recent high valuation
- •Why “stealth” can be a founder advantage: fewer external expectations while iterating
- •Operating focus vs. public messaging and its recruiting/branding trade-offs
- •Company culture emphasis: ambition and execution (“Attack.”)
- 1:58 – 4:14
Primer: what Applied Intuition builds and why vehicle intelligence is different
Qasar outlines Applied Intuition’s core mission: bringing AI into safety-critical, real-world vehicles across automotive and beyond. He frames the company as building “vehicle intelligence” with three major lines of business that grew from simulation and tooling into a full stack.
- •Vehicle intelligence across cars, trucks, tanks, aircraft, and more
- •Three pillars: engineering tools, vehicle OS, and autonomy/applications
- •Safety-critical requirements make validation/testing central
- •Product-first orientation vs. research projects that burn capital
- 4:14 – 6:10
The Microsoft analogy: tools → OS → apps for the vehicle ecosystem
They explain why the company expanded from tools into an operating system and then applications, drawing an explicit parallel to Microsoft’s historical progression. The key thesis: autonomy and next-gen vehicle software require deep integration with manufacturers rather than standalone “self-driving startups.”
- •Tooling improves when you also build applications that stress the stack
- •A vehicle OS is required to deploy and run advanced capabilities reliably
- •Manufacturers must be “in the loop” for autonomy to scale
- •Positioning: Tesla-like stack delivered via partnerships (no hardware)
- 6:10 – 6:56
Inside the Vehicle OS: embedded stack, updates, middleware, and cost reduction
Qasar breaks down the practical meaning of a “vehicle operating system,” from bootloaders through middleware and applications. The discussion highlights consolidation of many small ECUs into centralized compute, enabling both lower cost and richer functionality.
- •Full embedded stack: bootloader control, OS layer, middleware, apps
- •Reliable over-the-air updates require controlling low-level components
- •Centralizing compute reduces wiring harnesses and redundant modules
- •“Software replaces hardware” to add features while cutting bill of materials
- 6:56 – 11:07
Android lessons applied to vehicles: compatibility, hardware diversity, and safety constraints
Peter draws on Android’s success running uniformly across diverse hardware to explain how Applied Intuition approaches chipset/vehicle diversity. They contrast consumer computing abstractions with the realities of safety-critical embedded systems: longevity, reliability, and tight cost constraints.
- •Android’s key innovation: uniform apps across varied hardware with consistent UX
- •Compatibility Test Suite as a model for enforceable standards at scale
- •Embedded systems constraints: reliability, field lifetime, cost sensitivity
- •Achieving advanced capabilities within strict automotive-grade constraints
- 11:07 – 12:04
Who uses Applied Intuition: OEM partnerships and the Porsche example
Elad asks about customers, and Qasar shares a flagship public relationship with Porsche. They discuss why even top-performing OEMs face pressure to match Tesla and emerging Chinese competitors on software-driven experience.
- •Public “hero” customer: Porsche and why it’s an important validation
- •OEMs need parity with Tesla-like software experiences
- •Competitive pressure comes from both Tesla and China’s new entrants
- •Applied Intuition’s role: enabling modern vehicle software without OEM rebuilding everything alone
- 12:04 – 16:02
China’s EV/software surge: product quality, subsidies, and ecosystem dynamics
The conversation shifts to Chinese automakers (e.g., BYD, Xiaomi) and their rapid progress in autonomy and features. Qasar argues the products are genuinely strong—often better—while noting the role of state subsidies and the likelihood of consolidation into powerful platform suppliers.
- •Chinese vehicles: impressive autonomy and features; strong local competition vs. Tesla in China
- •Blank-slate advantage: fewer legacy platforms and constraints during the EV transition
- •State support/subsidies treat autos as strategic national infrastructure
- •Emergence of platform suppliers (e.g., Huawei’s automotive stack) enabling OEM focus on brand/manufacturing
- 16:02 – 20:49
Europe and industrial policy: tariffs, manufacturing localization, and “fight” mentality
Peter and Qasar discuss Europe’s vulnerability if it remains open to imports without building competitive local manufacturing and supply chains. Qasar argues automation reduces labor arbitrage, making local production more feasible, and warns against complacency in industrial strategy.
- •Trade balance and the long-run push to localize manufacturing
- •Europe “asleep at the wheel” vs. need for competitive urgency
- •Automation reduces labor-cost advantages; fully automated factories narrow gaps
- •Maintaining domestic manufacturing preserves know-how, jobs, and strategic resilience
- 20:49 – 21:51
Can robotics re-shore manufacturing? Founder opportunities in automated industry
Elad asks whether robotics and factory automation create a window to rebuild industrial capability. Qasar points to a founder-optimistic moment where venture capital is backing industrial and defense-adjacent manufacturing innovation, citing examples like Re:Build and Anduril.
- •Automation as an opportunity to make manufacturing location-agnostic and cost-competitive
- •Industrial innovation is increasingly venture-fundable (not just PE rollups)
- •Founder playbook: build in hard, real-world domains with capital support
- •Reshoring framed as both economic and capability/knowledge strategy
- 21:51 – 26:40
Training autonomy models: data collection, synthetic data, and technique convergence
Qasar explains Applied Intuition’s autonomy work across trucking, aerial, and maritime domains, emphasizing the difficulty of collecting real-world data. He describes multi-generation synthetic data approaches and how newer model techniques unlock value from older data corpora.
- •Autonomy programs: L4 trucking (notably Japan), plus aerial and maritime
- •Data is harder to collect outside on-road driving; formats and pipelines matter
- •Synthetic data evolution: from graphics-heavy simulation to diffusion/Gaussian splatting methods
- •Strategy: wait for ecosystem convergence on effective techniques, then scale execution
- 26:40 – 32:02
Safety and regulation: liability, media incentives, and verification/validation
They debate the safety bar for autonomous systems and why public perception often diverges from statistical reality. The discussion covers liability questions (driver vs. vehicle responsibility), sensational media coverage, and the need for provable safety via verification and validation.
- •Current systems can already exceed human safety on key metrics (e.g., disengagement, accidents per mile)
- •Liability framing drives policy: who is responsible when automation is active?
- •Media incentives amplify rare AV incidents vs. routine human-caused fatalities
- •Verification/validation work aims for provable safety, though perfection is impossible
- 32:02 – 36:44
Adoption timeline and monetization: 2025–2030 rollout dynamics and commoditization risk
Qasar predicts broad availability of FSD-like features across major US OEMs within five years, with Waymo expanding city by city. They warn of a major business risk: autonomy commoditizing before companies recoup massive R&D investments, contrasting Waymo’s open questions with Tesla’s clearer monetization.
- •US outlook: widespread FSD-like systems becoming common in ~5 years
- •Waymo expansion resembles early smartphone adoption: slow then sudden
- •Expectation shift: autonomy priced down toward “free” by 2030–2035 pressures
- •Monetization risk for high-capex autonomy efforts if commoditization arrives early
- 36:44 – 38:46
Autonomy reshapes cities: parking, mobility demand, and productivity impacts
They explore second-order effects of self-driving: urban layout, parking utilization, and how mobility could become on-demand. Qasar argues autonomy could increase productivity and reduce wasted city space, while also noting demand could rise as friction drops (like texting/calling did).
- •Cities are fundamentally designed around cars: parking, store placement, hospital access
- •On-demand fleets could reduce the need for prime-location parking lots
- •Lower travel friction can increase total miles traveled even if parking/traffic dynamics change
- •Optimistic view: reclaim urban space for parks and reduce fatalities simultaneously
- 38:46 – 42:08
AI beyond autonomy: developer tooling, in-cabin multimodal experiences, and defense swarms
Elad and Qasar discuss how AI changes vehicle software development and the user experience inside vehicles and industrial machines. They also describe defense shifting from “one operator, one machine” to “one operator, many machines,” requiring new interfaces for coordinating autonomous swarms.
- •AI-assisted software development for automotive-grade constraints (coding tools adapted to safety-critical needs)
- •In-cabin/industrial experiences: personalized, multimodal interaction with machines (e.g., mining equipment)
- •Operational safety gains from incremental intelligence in industrial settings
- •Defense evolution: collaborative autonomy, swarm coordination, comms/RF and data movement challenges
- 42:08 – 45:21
Design and hiring: building HMI for multimodal vehicles and scaling a deeply technical team
They close on Applied Intuition’s emphasis on design as end-to-end experience (not just pixels) and the unique HMI challenges of multimodal vehicles. Finally, they outline hiring priorities: heavy concentration of software engineers across systems, OS, and AI roles, reflecting a product (not services) company.
- •Design focus: multimodal vehicle interfaces, identity, personalization, and HMI depth
- •UX in vehicles involves multiple screens, sensors, and context—not just voice/chat
- •Hiring broadly (100+ roles), with strong demand for AI talent and systems engineers
- •Company identity: highly technical, product-driven, engineering-heavy organization
