In recent weeks, the AI boom that propelled the economy for the last two years has been shaken. Despite billions of dollars of investment, an MIT report found that 95% of AI pilots had no measurable impact on companies' bottom line. Even Sam Altman has suggested that investors are “overexcited” about AI.
While it’s hard to guess where we are in the hype cycle, it’s important to bear in mind that the long-term impact of AI is still in its infancy. From our perspective working with hundreds of Tough Tech companies, many of the ventures we see gaining momentum — with investors, partners, and potential customers — are those that build AI into the very fabric of their solutions.
I’m not talking about employing AI tools like ChatGPT to accelerate routine tasks and gain efficiencies. This is about the value a business is creating at its core: building a AI-native solutions that accelerate discovery, reduce costs, and create new value for customers. In particular, we’ve seen several successful (and potentially more fundable) companies at the intersections of AI and advanced materials discovery or life sciences platforms.
In light of recent headlines, investors may be more skeptical around AI claims than they were before. But they still recognize the potential value at stake, especially in the massive impact Tough Tech solutions pursue. To win them over, startups will need to demonstrably prove that their AI solutions can create value that was not possible before. In a difficult funding environment, it could make all the difference.
“From our perspective working with hundreds of Tough Tech companies, many of the ventures we see gaining momentum — with investors, partners, and potential customers — are those that build AI into the very fabric of their solutions.”
A Changing VC Lens: Speed, Automation, and Scalability
It’s no secret that raising capital for early-stage Tough Tech startups requires patience and a long-term vision. But capital providers are increasingly drawn to companies that signal faster iteration cycles, cost-aware scalability, and a defensible, platform-level advantage. Deep integration of AI can telegraph all three, and create value in many ways, including:
Speed to Market
AI-native companies can iterate and discover new solutions in days, not years, outpacing competitors and incumbents and achieving a first mover advantage. From an investor perspective, this could mean quicker proof-of-concept milestones and even reduced time to revenue.
Lower Operating Costs
Training and using AI at scale is not without its costs, from both a financial and energy consumption perspective. That said, AI-native companies are more likely to have a lighter cost structure and better contribution margins once initial training and R&D are complete. AI-driven approaches can also create a competitive moat against well-funded incumbents locked into expensive legacy approaches.
Commercial Scalability
AI-driven solutions can scale like software, even when grounded in hardware, biology, or materials in Tough Tech. The same algorithmic foundation used to discover new solutions for one domain can be rapidly applied across multiple product lines, market segments, or even adjacent industries without expensive new R&D cycles. Each new application leverages existing assets, allowing Tough Tech companies to achieve scalable growth profiles that investors expect.
AI-Native Companies in the Real World
The advantages of AI-native companies aren’t hypothetical, they’re playing out in some of the most compelling companies in the field. Below are a few examples of companies that have built AI into the core of their offering, including some Resident companies at The Engine.
Insilico Medicine
Insilico Medicine is the first company to bring a drug discovered and designed by generative AI into Phase II clinical trials with patients. The drug, INS018_055, is designed to treat Idiopathic pulmonary fibrosis (IPF), a rare and incurable lung disease that affects 5 million people worldwide and typically leads to death within 2-5 years of diagnosis.
Even if the clinical trials ultimately fail, the drug has already shown the promise of AI to accelerate drug discovery, cutting in half the time it traditionally takes to go from target discovery to Phase 1 clinical trials at a fraction of the cost.
Quantum Formatics
Much like how Insilico is applying AI to discover new drugs, Quantum Formatics is applying AI to discover new superconductor materials. These materials are the key to unlocking scalable fusion energy, lossless energy transmission, and practical quantum computing.
The company's end-to-end discovery platform combines generative AI, quantum mechanical simulations, and rapid experimentation, which Quantum Formatics claims can find candidate materials 10,000x faster than traditional methods, greatly accelerating the discovery process.
Phare Bio
Antibiotics have been hailed as the most important medical advance of the 20th century, saving untold millions of lives. But bacteria have aggressively evolved antimicrobial resistance (AMR) to these drugs, and the antibiotic discovery pipeline has been unable to keep up. By 2050, AMR is expected to kill 10 million annually, surpassing cancer as the leading cause of death.
Much like Insilico, Phare Bio is applying generative AI to accelerate the drug discovery pipeline, building the world’s first AI-driven antibiotics discovery engine, AIBiotics. Currently in Residency at The Engine, Phare Bio claims its platform can screen 1000x as many antibiotic candidates, 5x faster, at a 100x greater success rate, providing hope in the fight against AMR.
These companies don’t simply use AI; they are AI-native. The technology isn't a layer on top, it's what makes their entire solution possible.
AI as a Design Principle
We’re still in the early innings of what AI can unlock across Tough Tech domains. But I believe this much is clear: the companies that treat AI not as a tool, but as a design principle — one that guides what they build, how they build it, and how they scale — have an edge in fundraising as well as in operations.
This isn’t to say that AI integration is a prerequisite for fundraising; we have many teams in residence at The Engine and in our ecosystem that are successfully fundraising without it. And integrating AI as a design principle should never be forced: it doesn’t work for every business model and is far from the only way startups can create value. But it’s worth exploring, because in a difficult funding environment, it could make all the difference.
So here are a few questions I’d encourage Tough Tech teams to ask:
Where in our product or solution could AI create disproportionate value?
Are we collecting the kind of data that allows AI to become a true engine of discovery or performance?
Can we build a business model where AI not only improves outcomes, but reduces cost and improves scalability?
If we started from zero today, how would we design our product differently with AI as a core component? Put differently, if we were to start an AI-native competitor to ourselves, what would that look like?
The Tough Tech path will always be a demanding journey. And it’s worth it, for all of our benefit. But AI-native products can accelerate the path and create new leverage with investors and customers. In today’s venture climate, especially in light of federal research cuts, that leverage is creating a meaningful fundraising advantage for many Tough Tech teams.
