Scientific Breakthroughs That Drive Engineering Innovation

Big breakthrough: A few months my lab at MIT introduced SPARKS, our autonomous scientific discovery model. Since then we have demonstrated applicability to broad problem spaces across domains from proteins, bio-inspired materials to inorganic materials. SPARKS learns by doing, thinks by critiquing itself & creates knowledge through recursive interaction; not just with data, but with the physical & logical consequences of its own ideas. It closes the entire scientific loop - hypothesis generation, data retrieval, coding, simulation, critique, refinement, & detailed manuscript drafting - without prompts, manual tuning, or human oversight. SPARKS is fundamentally different from frontier models. While models like o3-pro and o3 deep research can produce summaries, they stop short of full discovery. SPARKS conducts the entire scientific process autonomously, generating & validating falsifiable hypotheses, interpreting results & refining its approach until a reproducible, fully validated evidence-based discovery emerges. This is the first time we've seen AI discover new science. SPARKS is orders of magnitude more capable than frontier models & even when comparing just the writing, SPARKS still outperforms: in our benchmark evaluation, it scored 1.6× higher than o3-pro and over 2.5× higher than o3 deep research - not because it writes more, but because it writes with purpose, grounded in original, validated compositional reasoning from start to finish. We benchmarked SPARKS on several case studies, where it uncovered two previously unknown protein design rules: 1⃣ Length-dependent mechanical crossover β-sheet-rich peptides outperform α-helices—but only once chains exceed ~80 amino acids. Below that, helices dominate. No prior systematic study had exposed this crossover, leaving protein designers without a quantitative rule for sizing sheet-rich materials. This discovery resolves a long-standing ambiguity in molecular design and provides a principle to guide the structural tuning of biomaterials and protein-based nanodevices based on mechanical strength. 2⃣ A stability “frustration zone” At intermediate lengths (~50- 70 residues) with balanced α/β content, peptide stability becomes highly variable. Sparks mapped this volatile region and explained its cause: competing folding nuclei and exposed edge strands that destabilize structure. This insight pinpoints a failure regime in protein design where instability arises not from randomness, but from well-defined physical constraints, giving designers new levers to avoid brittle configurations or engineer around them. This gives engineers and biologists a roadmap for avoiding stability traps in de novo design - especially when exploring hybrid motifs. Stay tuned for more updates & examples, papers and more details.

"A.I. hallucinations... are dreaming up riots of unrealities that help scientists track cancer, design drugs, invent medical devices, uncover weather phenomena and even win the Nobel Prize." A nice New York Times article "How Hallucinatory A.I. Helps Science Dream Up Big Breakthroughs" delves into the value of "hallucinations" in scientific advances. (Gift link in comments). Examples in the article include: 🌟 Nobel Prize recognition for "De novo protein design" David Baker’s groundbreaking work at the University of Washington has redefined what’s possible in protein engineering. Using AI hallucinations, his lab designed entirely new proteins from scratch—an achievement once considered "almost impossible." These proteins, numbering over 10 million, include innovations like cancer treatments and tools for combating viral infections. Baker’s work earned him the 2023 Nobel Prize in Chemistry. 🏥 Medical innovation with AI-designed catheters Anima Anandkumar and her team developed a novel catheter design using AI hallucinations to combat a major global health issue: urinary tract infections. Their model generated thousands of possible geometries before selecting one featuring sawtooth-like spikes lining the inner walls. These spikes prevent bacteria from adhering and traveling upstream to the bladder, drastically reducing bacterial contamination. The device is currently under discussion for commercialization. 💊 Accelerated drug discovery MIT professor James J. Collins is using AI to transform antibiotic discovery. By prompting models to dream up completely new molecular structures, his team can quickly identify promising drug candidates. This process, which used to take years, now takes just days, speeding up the fight against drug-resistant bacteria. Collins highlights hallucinations as a tool for sparking creativity in molecular design, a critical area for global health. 🌪️ Advances in weather forecasting Amy McGovern’s work at the University of Oklahoma shows how A.I. hallucinations can improve weather predictions. By generating thousands of probabilistic forecast variations, AI helps uncover hidden factors driving extreme weather events like heat waves. McGovern describes these AI outputs as invaluable for spotting unexpected patterns in the atmosphere. 🖼️ Sharpening medical imaging At Memorial Sloan Kettering Cancer Center, Harini Veeraraghavan has used AI hallucinations to improve medical imaging. By applying the technology to sharpen blurry MRI scans, her team enhances diagnostic accuracy. Their work, described as “hallucinated MRI,” has the potential to change how radiologists interpret scans, especially when clarity is crucial for finding abnormalities. The lesson: hallucinations are a feature, not a bug, if we understand their nature and use AI outputs appropriately.

AI’s biggest breakthroughs won’t come from chatbots—they will come from AI that models the physical world, not just language. While Large Language Models (LLMs) enhance efficiency in summarization, content generation, and automation, the real economic transformation is being driven by Large Quantitative Models (LQMs). Biopharma: LQMs simulate molecular interactions at the atomic level, accelerating drug discovery by evaluating millions of compounds virtually—long before clinical trials begin. Energy & Materials: AI is unlocking breakthroughs in battery chemistry, lightweight materials, and sustainability, replacing costly trial-and-error experimentation with precision-driven discovery. National Security & Navigation: AI-powered sensing systems are eliminating reliance on GPS in contested environments, delivering precise positioning even when satellite signals are compromised. Unlike LLMs, which rely on probabilities, LQMs are deterministic—rooted in physics, chemistry, and mathematics. They don’t just predict outcomes; they drive new scientific insights and create tangible economic value. The next wave of AI isn’t just about processing information—it’s about transforming industries at their core.

Concrete is the second most consumed material after water. But it has a deadly weakness: it cracks... These cracks let in water and oxygen that corrode steel reinforcement, threatening structural integrity. This is where self-healing concrete comes in - the biggest breakthrough in construction materials in decades. The secret? Bacteria. Scientists use Bacillus subtilis bacteria that can survive concrete's harsh alkaline environment. During manufacturing, bacterial spores and calcium nutrients are mixed directly into concrete. These remain dormant until a crack forms. Then the magic happens: When a crack forms, water and oxygen enter. This awakens the dormant bacteria, which consume embedded calcium lactate. As they metabolize this food, they produce limestone and naturally fill the crack. The process works automatically, with no human intervention. It's like your body healing a cut, you don't direct cells to close wounds, they just do it. The results are remarkable: At Delft University, researchers saw cracks repaired in just 60 days. Even more impressive: bacteria-treated concrete showed 40% higher strength after 7 days and 45% after 28 days versus traditional concrete. The implications are enormous: • Eliminates expensive repairs and reduces maintenance budgets • Could help improve America's C-grade infrastructure (ASCE rating) • Reduces environmental impact as less new concrete is needed • Fewer repairs mean reduced environmental disruption We're entering an era of living infrastructure, materials that respond to their environment. This convergence of biology and materials science is creating entirely new possibilities for how we build. Self-healing concrete isn't just an innovation, it's part of a fundamental shift in how we think about the structures we rely on every day.

US Develops Record-Breaking Armor Material with 100 Trillion Bonds Per Square Centimeter Northwestern University Scientists Create Breakthrough in Mechanically Interlocked Materials Researchers at Northwestern University have achieved a groundbreaking milestone by creating the strongest-ever armor material. With a staggering density of 100 trillion mechanical bonds per square centimeter, this two-dimensional material is set to redefine the future of lightweight, high-performance protective gear. Key Highlights • First-of-Its-Kind Material: This innovation is the world’s first two-dimensional mechanically interlocked material, combining exceptional strength and flexibility. • Origins of Mechanical Bonds: The concept of mechanical bonds, first introduced by Nobel laureate Fraser Stoddart in the 1980s, laid the foundation for this development. Stoddart’s work on molecular machines earned him the 2016 Nobel Prize in Chemistry. • Challenges Overcome: Previous attempts to integrate mechanically interlocked molecules into polymers were unsuccessful due to difficulties in forming medium-sized rings that could thread other molecules. How It Works • Mechanically Interlocked Molecules: The new material uses mechanically interlocked molecules arranged in a dense two-dimensional lattice. • Chemical Engineering Breakthrough: By solving the challenge of threading molecules through rings, researchers created a structure that maximizes bond density, achieving unprecedented toughness and flexibility. Applications and Impact 1. Advanced Body Armor: The lightweight and durable properties of this material make it ideal for next-generation protective gear. 2. High-Performance Materials: Beyond armor, the technology could be applied in aerospace, automotive industries, and infrastructure to create stronger yet lighter components. 3. Molecular Machines: This advancement further expands the scope of molecular machines, enabling new functionalities in nanotechnology and materials science. A Glimpse into the Future William Dichtel, a professor of chemistry at Northwestern University, emphasized the novelty of this breakthrough: • “These mechanically interlocked rings are the building blocks of a material that achieves strength without sacrificing flexibility,” Dichtel explained. The research team’s work is a testament to decades of progress in chemistry, bringing the vision of mechanically interlocked molecules from concept to reality. As this technology develops, it could redefine the materials industry, offering lightweight, high-strength solutions for a wide range of applications.

Tech isn't moving fast. It's EXPLODING. AI achieves 88% medical diagnosis accuracy. Quantum research promises ultra-fast charging. Lab-grown organs significantly reduce rejection risk. Here's what's converging to reshape our world: We're witnessing something remarkable. Not isolated breakthroughs, but technologies converging into interconnected transformation. Take Orion AI's o1-preview system. It just scored 88.6% accuracy diagnosing complex medical cases while human doctors achieved 35%. But here's what matters: it explains its reasoning step by step. We're not replacing doctors. We're augmenting their capabilities. Meanwhile, quantum battery research explores revolutionary possibilities. Scientists study quantum entanglement for ultra-fast energy transfer. Research suggests potential for batteries with decades-long lifespans and rapid charging. The end of constant charging could be approaching. The bioengineering revolution goes deeper. Labs are 3D printing functional organoids from patients' own cells, layer by layer. When you grow tissue from someone's DNA, rejection risk drops dramatically. Your body recognizes itself. But the convergence accelerates beyond healthcare. NAQI's neural earbuds just won CES 2025's Innovation Award by reading facial movements to control devices. No hands, no voice commands. Eye and facial movements become your interface. Accessibility technology shows the future of human-computer interaction. Intel's neuromorphic processors represent a paradigm shift. Their Loihi 2 chip processes information like your brain does - up to 10x more efficiently than GPUs for specific pattern recognition tasks. Hala Point packs 1.15 billion artificial neurons into a system that learns continuously while using minimal power. Transparent solar panels using quantum dots capture invisible light wavelengths. Early prototypes show buildings could generate significant power. Skyscrapers evolving into energy producers. Digital twins revolutionize how we plan and test. Singapore uses virtual city replicas to optimize traffic before implementing changes. Surgeons practice complex operations on digital models of their patients. Space tourism edges closer to reality. Pioneer Station targets 2026 for civilian zero-gravity experiences. Commercial space travel transitions from science fiction to near-term possibility. These aren't separate revolutions. AI accelerates bioengineering discoveries. Quantum computing enhances neural network optimization. Neuromorphic chips support real-time digital twins. Each breakthrough amplifies the others. Technology isn't just advancing. It's creating new possibilities where digital, biological, and quantum innovations increasingly intersect. The question isn't which technology wins. It's how they connect.

The future of scientific research just shifted into overdrive Researchers at North Carolina State University have developed an AI-powered autonomous laboratory that accelerates materials discovery by 1,000%, fundamentally changing how we approach some of humanity's most pressing challenges The breakthrough lies in replacing traditional steady-state experiments with dynamic flow experiments, where chemical mixtures are continuously varied through the system and monitored in real time Instead of capturing a single snapshot, this approach creates "a full movie of the reaction as it happens," generating 20 data points where conventional methods would produce just one This isn't just an incremental improvement, it's a paradigm shift The system's streaming-data approach enables machine learning algorithms to "make smarter, faster decisions, honing in on optimal materials and processes in a fraction of the time" The implications ripple across industries critical for our future: - Faster battery development for electric vehicles - Accelerated solar panel efficiency improvements - Rapid advancement in sustainable manufacturing materials. Professor Milad Abolhasani, who led this research, envisions a future where "scientists could discover breakthrough materials for clean energy, new electronics, or sustainable chemicals in days instead of years, using just a fraction of the materials and generating far less waste" We're witnessing AI evolve from analyzing existing data to actively conducting scientific research itself This autonomous lab represents the convergence of artificial intelligence with physical experimentation, a combination that could redefine the pace of innovation across multiple sectors The research, published in Nature Magazine Chemical Engineering, demonstrates that AI's greatest impact may not be in replacing human tasks, but in amplifying human capability to solve complex problems at unprecedented speed Read more about this breakthrough: https://lnkd.in/dkQKaD-d

🔬✨ Electrons, Enzymes & Algorithms – The New Toolbox of Chemistry When PCR revolutionized biology, it gave every scientist a must-have tool. But what’s the PCR of Chemical Synthesis? For decades, chemists worked like sculptors — flasks, hoods, heating mantles, and a handful of clever catalysts were enough to carve out life-saving drugs, energy materials, and specialty chemicals. But the last 10 years have changed everything. The modern chemistry lab is being reshaped by six powerful technologies: ⚡ Flow Chemistry – continuous, safe, scalable 💡 Photoredox Catalysis – light as a reagent 🔋 Electrochemistry – electrons as green reagents 🧬 Biocatalysis – evolving enzymes to evolve chemistry 🤖 AI Retrosynthesis – Corey’s vision → Synthia’s algorithms 🛠️ Automated Synthesis – toward self-driving labs These aren’t just upgrades. Together, they’re building a new ecosystem of discovery — faster, greener, smarter, and more convergent than ever before. 👉 In my new article, I dive into how each of these “Six Pillars of Modern Synthesis” is reshaping the chemist’s toolbox, supported by real-world breakthroughs and Nobel Prize stories. 💭 Which of these six do you think will transform your work the most in the next decade? #Chemistry #DrugDiscovery #Biocatalysis #Automation #Electrochemistry #Photoredox #AIinDrugdiscovery #FlowChemistry

To unlock the technologies of tomorrow, faculty, staff, and students in UW-Madison Department of Materials Science and Engineering engineering materials with revolutionary properties—from alloys that survive extreme environments, to medical devices powered by motion, to lithium-free batteries, to quantum materials that will reshape computing, imaging, and sensing. What stood out most is MS&E’s deeply collaborative culture. The department’s internationally recognized National Science Foundation (NSF) Materials Research and Science Center (#MRSEC) brings together industry partners and 30 faculty across UW–Madison to pursue fundamental discoveries and prepare the next generation of materials engineers. I was honored to explore some of these engineering breakthroughs firsthand: Additional Highlights • Dan Thoma demonstrated the game-changing potential of additive manufacturing and its ability to 3D-print alloys for everything from custom aviation components to advanced medical devices. • Chang-Beom Eom gave me a masterclass in thin-film epitaxy—showing how layering 2D quantum materials like transition metal oxides leads to novel electronic and magnetic behaviors. • Jason Kawasaki showed how stretching or straining ultra-thin crystalline materials unlocks new states relevant for quantum and superconducting computing. • Hyunseok Oh, a next-generation metallurgist, described his research on high-entropy alloys engineered to withstand extreme heat, perform in space, and improve recyclability. • In the Wisconsin Centers for Nanoscale Technology, Paul Voyles showcased cutting-edge imaging tools—advanced microscopy, spectroscopy, and x-ray analysis—that empower faculty and students to visualize materials at the atomic scale. • A visit to the MS&E undergraduate lab revealed major equipment upgrades and a bold, student-centered vision for transforming the undergraduate experience. • Dane Morgan discussed how AI is revolutionizing materials research—accelerating property prediction, atomic-level simulations, and data discovery by up to 100x. • He also shared his work with Adam Nelson and the Data Science Institute to launch the Wisconsin Undergraduate Research in Data Science (WISCURDS), the evolution of the Informatics Skunkworks. This initiative prepares undergraduates for data-driven research through real projects, teamwork, project management, and applied data science—skills shaping the future of engineering. The innovations I saw are foundational to our #EngineeringMoonshots and the future of UW-Madison College of Engineering’s leadership. Thank you to Izabela Szlufarska for your visionary leadership and unwavering commitment to our students. On, Wisconsin! #TheBadgerWay #EngineeringTheFuture #MaterialsRocks #MaterialScienceInnovation

For decades, materials innovation has split into two phases: Discovery (finding new compounds) and engineering (making them work in manufacturing). Discovery gets the funding and the headlines, (but that’s only 1% of the problem). The other 99%, engineering, is where decades are lost and billions are wasted. Lithium-ion batteries were invented in the 1980s, they didn’t make it into Teslas until 2010. Chemistry wasn’t the bottleneck, engineering was. Walk into any materials lab today and you’ll still see PhDs manually adjusting temperature, pressure, and chemical concentrations… then logging results into spreadsheets. They’re essentially running multidimensional optimization with hundreds of variables, one experiment at a time. It’s brilliant science… trapped in 1998 workflows. Now something new is happening. One example is Xiangyun "Ray" Lei, CEO of Ursa Simulation, Inc., who I met this summer, a founder driving the next wave of materials innovation. They are an AI-powered materials simulation engine that optimizes manufacturing processes for semiconductor, automotive, and chemical companies by finding the right combination of temperature, pressure, and material properties without manual experimentation. What used to take 100 physical experiments over six months can now be simulated as 10,000 virtual experiments over a weekend. 𖧹 No lab equipment. 𖧹 No rare materials. 𖧹 No months-long waits. Engineering time collapses from years to weeks. The semiconductor industry is already validating this shift. When TSMC sends specifications to Applied Materials or Lam Research, the billion-dollar question is simple: Who can figure out how to manufacture this fastest? Simulation capacity is now the competitive edge, and most teams can’t keep up. Some report turning away 90% of requests because they simply don’t have the compute. Imagine expanding that capacity tenfold. That’s not an incremental gain, that’s a new era of industrial speed, and it’s not just semiconductors. Automotive, chemicals, specialty materials, every field that depends on complex, interdependent systems face the same bottleneck. AI infrastructure, robotics, sustainability mandates, and GPU accessibility are all converging to make this possible right now. I believe materials simulation will be one of the defining opportunities of the next decade, combining massive market potential, high retention, and clear strategic value. Because the companies that master simulation won’t just accelerate innovation, they’ll own the infrastructure that powers the next industrial revolution. If you’re a founder building in materials simulation or engineering infrastructure, I’d love to hear what you’re seeing in the space.