One row per milestone; forecast the date each milestone is…: When will Discovery Loop first announce a commercial…

Part of: One row per milestone; forecast the date each milestone is first reached. If a milestone may plausibly never happen, express that through late percentiles. Shared background: Discovery Loop is a Delaware public benefit corporation announced August 5, 2026 by Jeff Dean, Sanjay Ghemawat, Oriol Vinyals, and Quoc Le, automating the full experimental loop of research (propose, run, evaluate, iterate), initially in ML research and engineering, acting as its own first customer, with stated applicability to the 14 NAE Grand Challenges. Seed round co-led by Radical Ventures and Khosla Ventures with Lightspeed, Kleiner Perkins, Doerr Capital, and Alphabet participating, terms undisclosed and unclosed at launch; Google Cloud supplies first-year compute, and Google announced a joint research framework with the company on ML systems. At launch: zero employees beyond the founders, no office, a single generic Member of Technical Staff posting in Palo Alto, no safety or policy or sales roles advertised, and a stated plan to hire a lean in-person founding team. Reference class: Google DeepMind's applied-science record shows fast wins where the experimental loop is digital with objective evaluation (AlphaFold at CASP, weather models, algorithm discovery) and slow or failed outcomes where the loop runs through the physical world or institutions (Streams shut down, disputed GNoME synthesis, Isomorphic reaching first trials only after five years). AlphaFold took roughly four years inside a fully staffed lab from project start to recognized breakthrough. Peer startups: Safe Superintelligence emphasized a small team and took nearly a year to reach roughly 20 employees and has shipped nothing public; Thinking Machines took about seven months to its first technical blog post and about eighteen months to a high-level safety document; Sakana AI took eight to twelve months to its first automated-discovery artifacts and saw its attribution claims disputed. — view all 4 rows

When will Discovery Loop first announce a commercial offering or paid partnership with a named external customer?✓#3

milestone_dateJan 15, 2029

10% by
Jul '27
25% by
Jan '28
median
Jan '29
75% by
Oct '30
90% by
Aug '33

rationale:

Status Quo and Commercial Posture Discovery Loop launched today (2026-08-05) as a strongly pre-commercial Delaware public benefit corporation. It has zero employees beyond its four founders, no office, and a stated roadmap of acting as its "own first customer" to automate its internal ML research and engineering stack before expanding to domain-specific NAE Grand Challenges 44 sources. The company's initial hiring footprint reflects this inward focus: the careers page lists a single generic Member of Technical Staff role with no business development, sales, or partnership functions jobs.ashbyhq.com. With elite venture backing and guaranteed first-year compute, the founders face no near-term revenue pressure to rush a commercial product to market.

The Google Exclusion Penalty The resolution criteria specifically exclude Alphabet's investment, Google Cloud compute, and the joint ML-systems research framework 3 sources. This structural constraint removes the single most likely initial paying counterparty. The company cannot trigger this milestone through an "easy" inter-company transfer or Alphabet-adjacent pilot; it must secure, and publicly name, a compensated engagement with a distinct external organization.

The Central Case A timeline of roughly 30 months from launch (early 2029) represents the most probable path to a named, paid commercial milestone. The closest structural analog is Isomorphic Labs—an Alphabet-adjacent, research-first entity targeting physical-world domains—which took approximately 26 months from its founding to announce compensated collaborations with Lilly and Novartis isomorphiclabs.com. Given Discovery Loop is starting from an absolute zero-to-one baseline without an inherited team, a slightly longer timeline is expected. While product-focused labs can move faster (Thinking Machines launched its Tinker API in roughly eight months reuters.com), Discovery Loop's mandate to automate the entire experimental loop requires substantial internal validation before an enterprise-grade platform or scientific partnership can be sold.

Acceleration Paths An earlier resolution—landing between mid-2027 and early 2028—could materialize through high-profile, founder-driven pilots rather than a mature product launch. The founders' unmatched networks could quickly yield a compute/hardware co-design collaboration, particularly given Jeff Dean's TPU background. Additionally, government funding for AI-for-science is abundant; the DOE's Genesis Mission recently awarded 278 projects, including phase-two company-led awards of up to $15 million, which could serve as a qualifying compensated research collaboration.

The Long Right Tail The late estimates stretch well into the 2030s to account for a meaningful probability that this milestone takes a decade or simply never occurs. Frontier-lab norms of secrecy often mean early revenue-generating deals remain unnamed, as seen with Periodic Labs research.contrary.com. Furthermore, the company could follow the path of Safe Superintelligence, functioning purely as a stealth research lab funded by continuous equity rounds techcrunch.com. There is also a non-trivial chance that Discovery Loop focuses entirely on monetizing its own proprietary IP (e.g., drug assets or chip designs) without ever taking on external platform customers, or that it is eventually quietly reabsorbed into Google DeepMind.

Accounting for the potential that early technical milestones could drive a rapid pivot into applied domains slightly accelerated our early-bound timeline for a commercial deal.