Ben LammWhy We're Living in a Biological Singularity With Ben Lamm | MOONSHOTS Live #297
In a Nutshell
Ben Lamm's Colossal Biosciences is engineering extinct species like woolly mammoths through precise gene editing, achieving 300+ simultaneous modifications with 90%+ efficiency. The company is developing artificial womb technology to deliver fully formed mammals outside surrogate mothers within 24 months, while building global biosafety infrastructure with T2T genome sequencing. Lamm positions this work at the intersection of synthetic biology, AI, and computing toward a "biological singularity" where organisms can be designed like software, with datasets being more valuable than AI models themselves.
These notes were generated by AI and may contain inaccuracies.
The conversation begins with a humorous reference to a backstage moment at the Abundance 360 conference in March, where Elon Musk was present and expressed interest in having a woolly mammoth as a pet. Colossal receives multiple requests for woolly mammoths, with the second most requested order being for a "teacup mammoth" - a miniature version small enough to fit in a handbag. The company has discussed engineering new melanin pigments to create pink variants, potentially appealing to figures like Paris Hilton, though this is not currently in development.
Ben Lamm founded four successful companies prior to Colossal in gaming, defense, artificial intelligence, and mobile phones, with no background in synthetic biology. He views this lack of domain expertise as a significant advantage, allowing him to enter rooms and ask questions without preconceived notions. Lamm emphasizes the importance of hiring people smarter than himself and surrounding himself with the best talent. He maintains that being curious and asking the right questions can lead to novel perspectives, noting that 10% of the time, experts tell him they "haven't thought about it that way before."
George Church is described as the "godfather of synthetic biology," serving as head of the genetics department at Harvard University. Church's lab has been highly productive, spawning many multi-billion dollar companies and being perhaps the most active biology lab globally in terms of startup formation. Church is noted for his collaborative nature, interest in open source approaches, and efforts to democratize genome sequencing technologies, reducing costs from billions to $100.
Lamm discusses the business model challenges in biology, particularly regarding foundational models. Colossal has worked with both OpenAI and Anthropic, gaining early access to advanced AI tools. For a significant period, these models served as "smart middleware" - useful for writing research papers but less effective for tasks like ancestral reconstruction or comparative genomics. The company has successfully used these models as infrastructure layers, avoiding the need to hire consulting firms like Deloitte or Accenture for system building.
Current AI models struggle with gene sequence inputs and fail to produce meaningful results from raw genomic data. However, Lamm emphasizes that datasets may be more valuable than models themselves, particularly for comparative genomics across evolutionary pathways. The company is working on what they call "Noah's Ark 2.0," proposing a global biosafety system with local repositories conducting telomere-to-telomere (T2T) sequencing.
Lamm argues that datasets hold greater value than models because they can be owned and controlled. The company has repeatedly emphasized this point. Large-scale comparative genomics across thousands of genomes provides insights that single-gene approaches cannot achieve, though complete genome insertion into AI systems won't produce perfect results.
The conversation addresses reviving extinct species, referencing Russian philosopher Nikolai Fyodorov and the late 19th-century philosophical movement called "Russian Cosmism," which argued that humanity's ultimate technological achievement would be resurrecting every human who ever lived. Colossal is positioned as the first company with a viable business model and technical path to potentially target the entire historical biosphere, extending beyond Fyodorov's vision to include non-human species.
Lamm clarifies that Colossal does not apply its technologies to humans, citing different investor requirements, governance needs, and regulatory processes. He received advice from Bob Nelson early on to avoid human applications until technologies reach stable stages, as FDA processes would create extended delays. Lamm notes that fewer than 100 Tyrannosaurus rex specimens have ever been found despite billions having existed, highlighting how few organisms leave fossil records.
Rather than simply recreating extinct species, Lamm believes engineering life from a programmable perspective will be more achievable. Environmental factors, epigenetics, and other variables mean that even successful cloning wouldn't guarantee identical outcomes - for example, cloning Mozart wouldn't necessarily produce someone who would "fix the terrible path music has taken."
Lamm confirms that with sufficient genetic predisposition understanding and environmental simulation, engineered organisms could potentially exceed original versions. The goal includes designing genomes that reflect specific phenotypes - creating larger animals, animals with longer snouts, or animals with wings. When asked about creating a Pikachu, Lamm acknowledges this as a popular request, noting greater public acceptance of genomic engineering for branded or fictional creatures than for human applications.
Lamm references Stewart Brand's famous quote about acting like gods, but prefers not to frame their work in those terms. The company evaluates species based on environmental contribution, food chain impact, extinction causes, and indigenous populations' spiritual connections. They consider educational benefits, pop culture appeal, and balance multiple factors without a perfect internal algorithm for decision-making.
Lamm describes being at the intersection of synthetic biology, computing, artificial intelligence, and eventually quantum computing as creating technologies more powerful than any weapon system. He believes mathematics is solved, physics is approaching resolution, followed by chemistry and biology, leading toward a biological singularity where complete knowledge bases enable organism design through computer-aided design software.
Lamm predicts that within 10 years, DNA synthesis capabilities will allow ordering custom organism designs, with technologies to engineer major phenotypes on basal organisms across biological branches becoming feasible. The company has progressed from celebrating 20 successful gene modifications to consistently achieving over 300 modifications with greater than 90% efficiency, now testing thousands of modifications.
The company has moved from 20 modifications to over 300 consistent modifications, testing 1,000 modifications currently. Lamm notes that basic amendments triple annually, though success rates vary. DNA synthesis is expected to replace multiple modifications more quickly, functioning as a machine generating desired genetic codes with higher success probability.
Colossal is developing artificial womb technology for mammals, with plans to deliver fully formed animal mammals outside the womb within 24 months, eliminating the need for surrogate mothers. Lamm distinguishes between technological feasibility and societal acceptance, ethics, and rights considerations for human applications.
The "iPhone moment" at Colossal would be when the world sees another extinct species brought back to life through precise gene editing. The demonstration of artificial evolution and pregnancy of mammals represents the second major turning point. Colossal is working on additional projects beyond the dire wolves that have not yet been shared publicly.
Protein folding has been resolved almost overnight and has become a goldmine of change in biotechnology. The equivalent problem of genotype-to-phenotype mapping involves predicting what a given DNA sequence will produce - whether that's a lemur, a dodo, or other species. The question is whether enough data can be collected to interpolate outcomes without needing to create each organism.
Colossal currently works on a species-based basis, attempting to extrapolate to large categories of animals. The size differential presents significant challenges - the Tasmanian tiger or thylacine involves a 1500-fold increase from a marsupial mouse to a marsupial wolf.
Understanding how size and other traits are organized in evolution, not just the genes themselves, but how and when they are organized and passed on, remains critical. While making a killer whale the size of a pet goldfish is probably not possible, expanding within certain levels of species may be feasible.
Dogs and certain groups of birds demonstrate tremendous size variation functions, exceeding 1500 times. From a non-trait engineering perspective focused on reviving extinct species, the goal is to look specifically at genes that led to particular traits. Separately, extrapolation based on biological classification allows for influencing size within 20 to 50 percent deviations in other species.
Coat color, size, and skin folds, scales, and feathers are all highly measurable and modifiable traits. Colossal maintains an AI team that works specifically on patterns, lines, snout length, and hair.
The woolly mice presented to the public required eight genetic modifications in the first generation. These consisted of eight base pair amendments delivered in a single connection process.
Moonshot projects can address one huge problem in one sector or be horizontal in addressing ordinary daily needs that intersect with multiple sectors. The reintroduction of extinct species is viewed as a systems issue, where the same systems modeling used to protect or restore species can be applied to human healthcare. Principles based on what is being built can be applied across multiple uses without being so narrow that they lose usefulness in other areas.
Mammals have emotional needs, and simply creating an animal does not absolve responsibility toward a living being. Colossal has a foundation that publishes all environmental conservation technologies as open source, making them freely available. They have 75 global partners and fund specific conservation projects.
Mammoths and elephants are highly social animals, so the goal is not to bring back single mammoths but rather herds. Colossal works on 16 different strains simultaneously. The target is tens of thousands of mammoths.
California condors and other near-extinction species have been bred and cared for with positive impacts on current populations. In Botswana, Elephant Havens works with orphaned elephants using AI and various tools to create artificial herds with matriarchal elephant community structures.
Research funded includes satellite imagery, drones, and artificial intelligence for elephant migration patterns and routes to understand social dynamics and hierarchy of movement. This technology and data affects elephant conservation work today, preventing orphaned elephants and enabling entire herd reintroductions into the wild.
George Church proposed that bringing back woolly mammoths could address global warming. Mammoths' natural habitat in Siberia and northern Canada consists of tundra, where they knock down trees. Without trees, grass grows, and grass sequesters approximately six times more carbon than trees while providing two to three times the albedo effect in reflecting light back into space.
With sufficient mammoth density in tundra regions, permafrost temperatures could be lowered by 6 to 8 degrees during summer months, limiting thawing. While some scientists at Colossal believe the impact may not reach that level, the mammoths provide net positive environmental benefits through ecosystem restoration regardless.
PhDs can be difficult to work with due to their models and stubbornness. When two PhDs disagree, the approach is to tell both they are right and help them move forward. George Church may be correct about mammoth density impacts at specific latitudes and longitudes, while others may be right that benefits exist for plants and animals without addressing carbon and climate issues. Either way, the net positive impact on elephants and the degraded tundra ecosystem remains beneficial.
Astromec is a multi-billion dollar company focused on foundational models and animal models that examine the whole tree of life. The company builds what are internally called "turning-behind models" that link to foundational models by understanding genome sequences, their evolution, timing, and reasons for not evolving in related groups.
These models examine climate triggers and build predictive capabilities to determine where traits originated and why they developed. Astromec's unique datasets about classification and understanding can correlate with foundational models like Mythos, acting as a competitor or directing focus when given a global Biovolt with millions of samples.
Half of global biodiversity may be lost in the next 25 years. This problem requires billions of dollars in federal funding across multiple governments working together, as zoos and nonprofits alone cannot solve it.
A backup infrastructure for life is needed, similar to how photos, text messages, and emails are backed up. Species loss creates negative ecosystem impacts and disrupts food chains. LifeBank USA stores every cell from the placenta when a baby is born, preserving the original "boot disk" containing stem cells, T cells, and natural killer cells.
A partnership with the U.S. Fish and Wildlife Service was announced in Dubai as part of Secretary Zinke's directive on backing up natural resources. Additional government partnerships remain unannounced. The framework is completely open source, welcoming nonprofits, academic institutions, corporations, and individuals to participate.
Curiosity drives learning new things and admitting what is not known. In an AI world where answers are readily available, maintaining childlike wonder and seeking answers remains valuable. People will help when asked - nine times out of ten, humanity responds positively to requests for assistance.
Senior advisors at Colossal were contacted through simple emails stating lack of understanding in specific areas, resulting in meetings with world experts who had no obligation to respond but chose to help.
Anthropic's announcement of wet labs creating something resembling CRISPR represents a pivotal moment showing AI companies understand the scientific background in biology. This confirms that healthier families and longer, healthier lives represent acceptable use cases for the technology.
An optimistic outlook prevails despite acknowledging scary moments ahead. Human ingenuity will overcome problems, though dialogue remains important. Media may be exaggerating current concerns, but safe outcomes are expected, similar to airplane turbulence where everyone lands safely.
Keep Peter H. Diamandis in your library
Save the videos and channels worth coming back to, and find them again in one place.
