Sign in to view Salil’s full profile
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Sign in to view Salil’s full profile
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Palo Alto, California, United States
Sign in to view Salil’s full profile
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
9K followers
500+ connections
Sign in to view Salil’s full profile
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
View mutual connections with Salil
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
View mutual connections with Salil
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Sign in to view Salil’s full profile
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
About
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
Articles by Salil
-
Hidden Road acquired by Ripple
Hidden Road acquired by Ripple
Uncorrelated portfolio company Hidden Road, a crypto prime broker that institutional investors and traders rely on for…
127
17 Comments -
Hydrolix: Can Non-AI Companies Go From $1M to $40M ARR in 2 Years?Apr 3, 2025
Hydrolix: Can Non-AI Companies Go From $1M to $40M ARR in 2 Years?
I can't usually talk about portfolio companies' revenue traction. But since the figure was already reported in the news…
274
13 Comments -
Data-Center Startup Crusoe Raises $600 MillionDec 13, 2024
Data-Center Startup Crusoe Raises $600 Million
Data-center startup Crusoe, where I initially invested $1M at $15M post in 2019 while at Bain Capital Ventures, raised…
148
14 Comments -
Salvatore Sanfilippo Returns to RedisDec 10, 2024
Salvatore Sanfilippo Returns to Redis
I’m very excited that Salvatore Sanfilippo, the creator of Redis, where I’m still on the board twelve years after…
193
1 Comment -
Astranis & others to Build Backup GPS Network for US Space ForceNov 4, 2024
Astranis & others to Build Backup GPS Network for US Space Force
Astranis (#uncorrelated) and three other companies (Axient, L3 Harris, Sierra Space) are being asked by the US Space…
74
3 Comments -
AstroForge Raises $40M to Mine Asteroids for PlatinumAug 21, 2024
AstroForge Raises $40M to Mine Asteroids for Platinum
AstroForge (portfolio) raised $40M to finish building a spacecraft in the next few months to rideshare on the Intuitive…
307
16 Comments -
My Downgrade from Tesla Model XMar 19, 2019
My Downgrade from Tesla Model X
I had shared this story with my partners while we were discussing potential investments in self-driving technology…
50
18 Comments -
Lumina: Crypto Portfolio Management & Trade Execution for Institutional InvestorsOct 2, 2018
Lumina: Crypto Portfolio Management & Trade Execution for Institutional Investors
We’re thrilled to announce Bain Capital Ventures’ investment in Lumina’s $4 million seed round. Lumina combines crypto…
21
2 Comments -
Commons Clause Stops Open Source AbuseSep 7, 2018
Commons Clause Stops Open Source Abuse
This article was published on TechCrunch on September 7, 2018. There’s a dark cloud on the horizon.
28
3 Comments -
How Nutanix acquired FrameAug 24, 2018
How Nutanix acquired Frame
A few years ago I discovered Frame, which lets you run any application in the cloud, in any browser, without any…
282
13 Comments
Activity
9K followers
-
Salil Deshpande shared thisI hadn't heard of Eigen AI. MIT PhD recent-grad AI whiz-kids, how would I know them? Fortunately MIT prof Matt Rhodes-Kropf (Tectonic Ventures), a dear friend and #Uncorrelated LP said, “Hey, you should take a look a this super cool seed deal I’m doing.” I love it when people say that. I committed after the first meeting. Six months later they get bought for $643M. Kudos to Ryan Hanrui Wang and the amazing team, and thank you Matt!Salil Deshpande shared thisToday, we're announcing that Eigen AI is joining Nebius (NASDAQ: NBIS). From day one, our mission has been Artificial Efficient Intelligence — building the world's most efficient engines for generating intelligence. With Nebius, we're working toward the best AI cloud, uniting Eigen's industry-leading AI model customization and inference software, with Nebius's global hardware and infrastructure footprint, so any developer or enterprise can run the best models at the best price, with no capacity ceiling. After close, Eigen's optimization stack will be integrated directly into Nebius Token Factory. The entire Eigen AI team is joining Nebius in full, establishing Nebius's engineering and research presence in the San Francisco Bay Area. To our customers, our team, our investors at Tectonic Ventures, E14 Fund, Uncorrelated Ventures, and AGI House Ventures, our angel investors, advisors, mentors, and supporters — and to the Nebius team for the conviction and partnership — thank you. Ryan Hanrui Wang, co-founder and CEO of Eigen AI, said: “We’re proud to join Nebius and work alongside the Token Factory team to push the boundaries of inference performance. Nebius has built a world-class AI cloud with a deep engineering culture that perfectly aligns with our own. Together, we are removing the friction of AI model customization and deployment so developers can run models reliably in production without managing the underlying infrastructure.” Full announcement at: https://lnkd.in/gqpJKBvH
-
Salil Deshpande shared thisGlad that #Uncorrelated was able to join Bessemer Venture Partners in preempting a $20M round in HockeyStack. https://lnkd.in/gTfP67ib HockeyStack’s key thesis was that Salesforce's data model was too restrictive, so much so that sales and marketing could never properly be unified. And since Salesforce educated the industry on what a CRM should be, all its competitors, such as HubSpot, copied their data modeling errors; thus it’s impossible to fully unify sales and marketing using those tools (e.g., for full end-to-end attribution).Salil Deshpande shared thisWe raised $50M to build the World’s First AI Revenue Agent: It runs New Business, Expansion, and Prospecting 24/7. AI is changing the world, but Sales teams still run on tribal knowledge, gut calls, and fictional CRM stages: - Your best rep closes 2-3x more than your median. Nobody knows why. - Pipeline reviews are guesswork, and forecasts break every quarter. - Every tool tells you what happened. None of them close you more business. HockeyStack’s Revenue Agents fix this in 3 key ways: 1. Build your winning Blueprint: HockeyStack reverse-engineers every deal you've ever won or lost and extracts the exact path to win, by motion and segment. This becomes the brain for Revenue Agents. 2. Clone your Top reps: Your top performers run plays that live in their heads. HockeyStack finds and deploys them across your entire team. 3. Deploy Revenue Agents: Dedicated Agents monitor every deal and account, execute the right moves autonomously, and flag risks. A Revenue Agent spots your $550K deal about to stall, identifies a new champion, gets the intro, and preps your rep. All before morning coffee. 300+ Enterprise companies run revenue on HockeyStack, including Microsoft, 8x8, and Yext. On average, teams close 48% more deals using the platform. If you have 30+ reps, HockeyStack will 2x their output. If you don't see it in a pilot, I'll donate $10K to a charity of your choice. Book a demo here: https://lnkd.in/gpDVQ4Wm Everything in my life has led to this moment: I grew up middle-class in Turkey, made $350K when I was 18 with my first SaaS, and dropped out of college after 3 months to build a generational company. Today, we are an 8-figure business. Fortune 100 companies trust and love us, and 90+ of the smartest people in tech joined our team last year to build the Agent that Revenue Teams deserve. Thank you to: Our investors who tripled down (Bessemer Venture Partners, Y Combinator, Alexandra Sukin, Jeremy Levine, and Salil Deshpande), customers, team, and my co-founders.
-
Salil Deshpande shared thisExcited to receive OpenAI stock in the all-stock acquisition of Promptfoo (#uncorrelated). Much gratitude to George Mathew, Ian W., Ganesh Bell for allowing Uncorrelated to invest early. https://lnkd.in/gTpaBXmUOpenAI acquires Promptfoo to secure its AI agents | TechCrunchOpenAI acquires Promptfoo to secure its AI agents | TechCrunch
-
Salil Deshpande shared thisThe code reviewer from CodeAnt AI (#uncorrelated) found 87 zero-day vulnerabilities across npm, PyPI, Maven, and NuGet. This is #1, the most rare and severe: a CVSS 10.0 vulnerability (the highest possible severity score), a critical authentication bypass hidden for six years in pac4j-jwt, a widely used Java security library.Salil Deshpande shared thisCodeAnt AI’s code reviewer just found a CVSS 10.0 vulnerability. CVE-2026-29000. The highest possible severity score. The vulnerability had been sitting in pac4j, one of the most widely used Java authentication libraries, for nearly 6 years. No static scanner caught it. No human reviewer caught it. Our AI code reviewer did. This was a complete authentication bypass caused by a logic gap between the JWT specification and the implementation in `pac4j-jwt`. What does that mean in practice? An attacker can craft a malicious authentication token and log in as any user, including an admin. - No password - No credentials - No stolen session The only thing they need is your server’s RSA public key, which is publicly available by design. The root cause: the library accepts a class of token that the JWT spec technically allows, but that should never bypass signature verification. When pac4j receives that token, it skips signature verification entirely and creates an authenticated session from attacker-controlled claims. We’ve shared the full technical write-up and working PoC in the links below. We don’t publish random code review benchmark claims. We publish CVEs. This is one of 87 vulnerabilities we’ve uncovered so far, including issues affecting packages with 500M+ weekly downloads, many of which are currently in responsible disclosure with maintainers. Kudos to Patrick Garrity and team at VulnCheck to expediting the CVE assignment process.
-
Salil Deshpande reposted thisSalil Deshpande reposted thisYour AI agent just wrote a Terraform change that touches 6 services. It has no idea. This is the reality most platform and infra teams are walking into. Agents have no awareness of cross-service dependencies, compliance requirements, or the blast radius of what they're proposing. They're generating change at a volume and speed that existing review processes were never designed to handle. Cory O'Daniel will sit down with special guest Kelsey Hightower to talk about what's actually broken in how most teams are approaching agentic infrastructure today: • Agents generating change at speeds humans can't review • Zero awareness of blast radius or compliance requirements • The validation layers that have to exist before agentic infra is safe to ship If your team is responsible for keeping production stable, this one's for you. 𝗛𝗼𝘄 𝘁𝗼 𝗷𝗼𝗶𝗻: Send 1–2 questions you want Cory and Kelsey to answer in the chat below, or email them to Luis@massdriver.cloud — we'll send you the link to register. Hope to see you there!
-
Salil Deshpande shared thisAs AI workloads eat up global computing supply, DRAM prices are surging like never before—creating major challenges for teams trying to plan out their data center spend. One of my portfolio companies, Mext, is addressing the largest cost component in the datacenter: server memory (DRAM). From being on the board of Redis for the last twelve years, I’ve learned a lot about the various aspects of this problem and approaches to solving them. Mext found a breakthrough, using new AI techniques, for dramatically reducing the amount of server-DRAM required to run applications, all while maintaining performance. It intelligently manages the server’s memory, keeping hot pages (i.e., those in use by applications) in fast DRAM and offloading cold pages (i.e., those less used) to a much less expensive memory tier (e.g., NVMe Flash). Key to the approach is ensuring cold memory pages that are about to be accessed by an application are back in DRAM before the application needs them or notices that they were gone. This is done without modifications to the application or the OS – so it can run in the cloud or on-premise. Mext thus allows applications to either run using less DRAM or keep their DRAM footprint but do more with it. They're hosting a webinar along with Fred Weber (former CTO of AMD) on Jan 22nd at 10am pacific time. You can register here: https://lnkd.in/gXe9GCeEWelcome! You are invited to join a webinar: RAM Prices Are Surging—What's Happening and What To Do. After registering, you will receive a confirmation email about joining the webinar.Welcome! You are invited to join a webinar: RAM Prices Are Surging—What's Happening and What To Do. After registering, you will receive a confirmation email about joining the webinar.
-
Salil Deshpande reposted thisSalil Deshpande reposted thisThe 4th Session of the "Stanford University #AI and #Investment Series", co-hosted by AIx2 and Stanford BASES brought together an exceptional group of investors with unique investment practices for a thought-provoking conversation. A huge thank you to our panelists for sharing such depth of insight. My key takeaways: 🔹 Salil Deshpande, General Partner at Uncorrelated Ventures — Second-largest solo GP — delivered a masterclass on reducing complex investment processes to first principles, and how he’s taking efficiency to the next level with AI, all while running a $750M fund with #zero_employees. Exceptional clarity and precision. 🔹 Saam Motamedi, Partner at Greylock Partners (established 1965), shared powerful perspectives on horizontal AI platforms and how #Anthropic —a Greylock portfolio company—is building #vertical_AI solutions. He also highlighted Greylock’s small-batch, hands-on investment model and their internally developed AI tools. 🔹 John Whaley, Stanford LLM Lecturer, 3x Founder, and Inception Studio Partner, offered candid insights on the current AI bubble, why smaller VCs may hold a competitive edge, and the rising influence of experienced founders in this new cycle. 🔹 Will Ballard, Chief Data Officer, Prime Unicorn Index & Cancer Focus Fund, walked us through a data-driven perspective on U.S. private markets and his work developing the U.S. Unicorn Index and pioneering it for #data_driven_investment. If you are an investor, reach out to us for access to the event #video_recording. And stay tuned for more speakers and upcoming sessions in the Stanford AI and Investment Series. Special thanks to Stanford BASES team, the largest entrepreneurship group in Stanford , Matthew Musson, Taarush Grover, Amir Zeinali, and Armita Hosseini for making this great event happen! #AI #Investment #Stanford #VentureCapital #PrivateEquity #PrivateMarkets #AIx2
-
Salil Deshpande reposted thisSalil Deshpande reposted thisCongratulations to Uncorrelated Ventures for winning both Emerging Performance of the Year: Venture Capital and Performance of the Year: Venture Capital under $500 million categories at the Private Equity Wire® US Awards 2025. See the full list of winners on our website. #PEWAwards Salil Deshpande
-
Salil Deshpande reposted thisSalil Deshpande reposted thisThis VC manages $750M across 3 funds with zero employees, partners, analysts, or even an EA while delivering 5-6x returns consistently. Meet Salil Deshpande, the second-largest truly solo GP by AUM. With 100+ infrastructure software investments and 300 LP relationships, Salil co-invests alongside Accel, Benchmark, Bessemer, and a16z without competing for control. Before going solo, Salil spent 14 years at larger firms, including Bain Capital, where he invested in 42 companies in 7 years (holding a record of the highest # of deals per year). His career track record sits at 6x TVPI with 35% IRRs. His deep expertise in infrastructure software enables him to evaluate technical depth with an expert lens, having seen hundreds of similar companies. With this, Salil built the foundation for a pull-based fundraising strategy: He doesn’t have to chase LPs. He simply sends a deck, gets on a one-hour Zoom, answers questions, sends materials, and then never follows up. This strategy delivers a 75%+ conversion rate across 100+ conversations. Today, Salil manages 300 LPs by himself by issuing quarterly letters of 5-15 pages that provide full context for every markdown, down round, and value decrease. This eliminates LP anxiety and reduces the need for catch-up calls, letting him spend 70% of his time on founders instead of investor relations. With more time for actual investing, he built his unconventional portfolio approach: Instead of fighting for 20% ownership stakes, Salil invests smaller checks of a couple $M each in 4x the number of companies, quadrupling his probability of capturing a decacorn. He builds positions over 3-4 rounds: say $2M at $25-50M post, then $3-4M at $100M post, then $3-4M at $250M post. This gets $10-20M invested at a blended $100-150M entry cost, delivering 40-60x returns at $10B valuations despite never owning 20% at any single round. He also has a true market-based partnership model, where he presents deals to 5+ domain experts at competing firms who invest only when the deals genuinely make sense, thereby eliminating politics and social obligation of traditional VC partnerships within the same firm. This delivers pure market signals with a 30% opt-in rate among investors. What surprised me most was Salil's operational efficiency, without even an EA. He outsources everything non-core, like fund administration, legal, and accounting (spends millions to buy back mental space). The freed-up bandwidth allows him to focus purely on technical evaluation. Salil's model works because of his established track record, narrow focus, and comfort operating against consensus when others pass on deals he still believes in. Salil’s systems to manage 300 LPs and evaluate 200+ deals annually are learnable. I broke down his entire operating model and his unique strategy that generates 6x TVPI with 35% IRRs returns in my latest deep dive. Share this post and subscribe to my Substack to get access. 👇 Link in first comment
-
Salil Deshpande reacted on thisSalil Deshpande reacted on thisI’m glad to report that Anthropic’s infamous blog post has been corrected, and Hiive’s name has been removed. The inaccurate post caused needless confusion amongst the investment community and unfair damage to Hiive’s reputation. Let me be clear: every investment through Hiive (in Anthropic or otherwise) is legitimate. They are authorized and approved whenever such authorization is required. There is no basis for anyone to suggest that they can be invalidated. Investors that have gained exposure to Anthropic through Hiive can be confident that they will participate in Anthropic’s success. Hiive’s mission is to build state-of-the-art liquidity infrastructure for private companies, their shareholders, and their investors. We are succeeding in that mission and have earned the trust of countless private companies. In fact, we manage liquidity exclusively for some of the world’s most valuable late-stage private companies. Had Anthropic approached us before their aggressive new stance and corresponding public statements (they did not), we would have gladly worked with them to deliver a unified message to the market. What this whole situation highlights is a structural flaw in the private market that has nothing to do with Hiive. There is an entire industry of insiders peddling access to valuable private companies in exchange for massive “pre-paid management fees” that are essentially brokerage fees. However, these firms are not licensed as brokers and are not subject to the rigorous compliance regimes of FINRA member firms like Hiive. Meanwhile, the FINRA member firms and platforms that are actually licensed to carry out brokerage activities and that meet the substantial regulatory burdens of doing so are the ones met with suspicion. It’s the product of decades of traditional practices and entrenched interests that have every reason to stop us. We will march forward in our mission undaunted. When truth is on your side, you have nothing to fear. #liquidity
-
Salil Deshpande reacted on thisSalil Deshpande reacted on this90% of today's software is embarrassingly easy to breach. We ran AI pentests on 200+ companies. Most were leaking PII, PHI, or other critical data. Annual pentest miss these. SAST & SCA tools with 10,000 findings bury the actual exploit. We need continuous AI offense, fused with codebase, network, & cloud context. That's what CodeAnt AI does. Get a free pentest today --> https://app.codeant.ai Pay only if we find high or critical issue.
-
Salil Deshpande reacted on thisETF Partners - The Environmental Technologies Fund
ETF Partners - The Environmental Technologies Fund
1wSalil Deshpande reacted on thisFor 5,000 years, project management has been governed by the “Iron Law” of time, cost, and quality. But today, the world is introducing a critical fourth element: carbon. We are looking back at our insightful conversation with Greg Lawton, CEO of our portfolio company Nodes & Links. Greg explains how Nodes & Links is utilising advanced analytics and AI to automate carbon calculations in real time. Rather than relying on historical accounting, project managers can now orchestrate complex construction systems to see the immediate carbon and financial impacts of their day-to-day decisions. The best part? Sustainability and efficiency go hand in hand. By predicting delays and reducing system inefficiencies, projects can finish months ahead of schedule which significantly lowers both project emissions and capital costs. Watch the full interview to learn how AI is helping the construction industry execute predictably, sustainably, and at pace: https://lnkd.in/eBJhrkPF #Sustainability #SustainabilityThroughInnovation #ProjectManagement #AI #ConstructionInnovation #VentureCapital #VCETF Partners Annual Investor Meeting: Interview with Greg Lawton (Nodes & Links)ETF Partners Annual Investor Meeting: Interview with Greg Lawton (Nodes & Links)
Experience & Education
-
Uncorrelated Ventures
******* *******
-
******** **********
**** ********** *********** undefined
-
-
******* **********
**** ********** ***********
-
View Salil’s full experience
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Welcome back
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
New to LinkedIn? Join now
Recommendations received
11 people have recommended Salil
Join now to viewView Salil’s full profile
-
See who you know in common
-
Get introduced
-
Contact Salil directly
Other similar profiles
Explore more posts
-
Matt Wachter
Barber National Institute • 4K followers
I’m lucky to work alongside colleagues like Richard Fruehauf, who constantly push my thinking. His latest piece unpacks why U.S. robotics and drone manufacturing faces uphill economics, and how public policy can help turn the tide. One key takeaway: 232 tariffs on robotics imports could help reset the economics that are holding back U.S. investment and production. It’s a sharp perspective on how policy can shape the future of advanced tech here at home. Well worth a read: https://lnkd.in/edMPTQPQ Richard Fruehauf, Robert J. Szczerba, Michael Lutzky, Edward Ovando, Karl Sanchack, Herman Herman, Jeff Legault, Phil English, Richard Fruehauf, Ryan Aument, Michael Guido, Sheila Fitzgerald Sterrett, David McCormick
19
2 Comments -
Bilal Zuberi
Red Glass Ventures • 65K followers
Chatting with ZEDEDA CEO, Said Ouissal, and he has an elegant way of explaining the evolution of industrial Edge. Zededa provides OS for orchestrating Edge AI. 2010s were about IoT/Edge Connectivity. 2020s were about Edge computing/processing. 2030s will be all about Edge AI.
36
4 Comments -
Naresh Soni
South Coast Angel Group - II • 7K followers
We’re excited to share that our fund has invested in Cerebras Systems. Artificial intelligence is entering a phase where infrastructure matters as much as algorithms. As AI adoption accelerates across industries, the demand for scalable compute to train and run increasingly complex models is growing exponentially. Cerebras is building technology designed specifically for this new era of AI. Their breakthrough approach centers on wafer-scale computing — a radically different architecture built to handle large-scale AI workloads with greater efficiency and speed. Instead of stitching together thousands of smaller processors, Cerebras developed a single, massive AI processor designed to simplify how large models are trained and deployed. What impressed us most about Cerebras is their bold rethinking of AI compute architecture and their focus on solving one of the most important challenges in the AI ecosystem: enabling organizations to run large AI models faster and at scale. As AI moves from experimentation to mission-critical enterprise deployment, the infrastructure powering these systems will become increasingly important. We believe Cerebras is well positioned to play a meaningful role in shaping that future. We’re excited to support the team as they continue to push the boundaries of AI infrastructure. #AI #ArtificialIntelligence #DeepTech #Infrastructure #Investing
26
3 Comments -
Lars Albright
5K followers
Excited to share that Miravoice has raised a $6.3M seed round — and we’re proud to partner with Nishant Jain, Danny D. Leybzon, Shreyas Tirumala, along with the whole Miravoice team. From the beginning, what drew Unusual Ventures to Miravoice was how clearly they saw a problem most organizations have simply learned to live with: collecting structured data through voice conversations is still manual, expensive, and slow. Despite all the progress in AI, large-scale phone surveys and interviews are still largely powered by call centers. Earlier voice AI solutions struggled with reliability and accuracy, making them difficult to trust for high-stakes data collection. Miravoice takes a fundamentally different approach. Read our full post below: https://lnkd.in/e8yJEeDS
131
6 Comments -
Hemant Mohapatra
Lightspeed India Partners… • 55K followers
Global defense landscape is changing rapidly with drones. Today all drone attacks happen "surface --> air --> surface" - drone takes from from land, moves in the air, and land back. They move "horizontally", can be sweep-detected via RF / radar or optical sensors, and thwarted via proximity fused ammo, or other means. Our company Kalam Labs has built state of the art surveillance & defense / attack drone infra that picks up a fleet of drones to stratosphere/near-space (100K feet) and conducts missions while dropping "vertically". Size of a eagle, flying higher than a commercial plane, dropping from stratosphere: practically impossible to detect & thwart. This is the sort of sovereign capability India needs and we are super excited to see Ahmad Faraaz Sashakt Tripathi & Harshit Awasthi building it!
34
-
Evan Nisselson
LDV Capital • 6K followers
Jazzed for this keynote by Jeff Erhardt at our 12th annual LDV Capital Vision Summit: The Materials Innovation Gap: Why AI Predictions Aren't Enough — And What It Will Take to Transform a $6 Trillion Industry RSVP: https://lnkd.in/eeUHN_m AI is transforming industry after industry, and generating enormous excitement in the world of materials science. Machine learning models can now predict millions of new material candidates with unprecedented speed and accuracy. At the same time, programmable material classes, including those recognized by the 2025 The Nobel Prize in Chemistry, could unlock breakthroughs in some of the most pressing challenges facing civilization, from environmental remediation to the energy transition. But their impact is constrained by something prediction alone cannot resolve: new materials must be physically produced, scaled, integrated, and qualified within the larger systems into which they are deployed, a journey that remains extraordinarily long, expensive, and unpredictable. Jeff will argue that closing this materials innovation gap requires treating AI and physical experimentation as equal partners, and building a new model of collaborative development that connects materials innovators and industrial partners far earlier in the discovery process.
13
-
Steve Vassallo
Foundation Capital • 16K followers
Cerebras is now a generational AI company. But between 2018 and 2019, one tiny component nearly killed the company. Back then we were deep into building the first wafer-scale system. Everything depended on a single part most people will never notice. The interposer. It sits between the main board and the dinner-plate-sized silicon wafer. It has to carry enormous power and data loads while having to survive massive thermal cycling. Cold to hot. Off to fully lit. Over and over again. We had just built the largest chip in the world. There was no existing supply chain for it so we had to single source this part. That alone is terrifying. One vendor, no backup. If it failed, the system failed. If the system failed, the company could have failed. For almost three years, this thing was the boogeyman in every board meeting. Would it maintain electrical continuity? Would it warp under load? Would it survive burn-in? Would it work at scale? For long stretches, the honest answer was “we don’t know.” We were burning millions of dollars per month. Time was not our friend. A redesign could have pushed us far enough back that the market might have moved on. At one point, we pulled in a group that I first worked with at IDEO and that used to be called "Failure Analysis". That is not who you call when things are going well! We went back to first principles. Physics. Materials science. Coefficients of thermal expansion. Power dissipation per square millimeter. Analyzing what happens when you push this much power through a system at these temperatures. Again and again. Slowly, painfully, we got there. Finally the interposer worked. Every company pushing the frontier runs into problems like this. There are rarely quick, clever fixes. You just have to do real failure analysis, over and over, until the answer starts to reveal itself. The Cerebras team is itself built from the strongest, toughest material I’ve seen. They earned every inch of the other side of that problem.
337
26 Comments -
Matt Logan
Earthshot Ventures • 5K followers
Varun Sivaram has a truly unique ability to pull together stakeholders across AI, energy, and policy. This is exactly why Earthshot Ventures invested in Emerald AI - there is a grand bargain to be struck in getting data centers online fast if they're willing to bring flexibility to the grid, and Varun is the man to strike that bargain. The latest illustration - bringing together $5T in market cap consisting of the world's leading chip company and the 5 leading IPPs in the US to make this happen!
17
1 Comment -
Yaniv Golan
lool ventures • 5K followers
AI is changing what “scale” actually means. In CTech by Calcalist’s 2026 VC Survey, Haim Bachar and Lee Ben-Gal share how we’re seeing the scaling metric shift - from headcount growth to output per person. This isn’t about doing more with less. It’s about building companies differently from day one - and rethinking how success is measured. See the full article below 👇
13
-
Irakli Kashibadze
University of California… • 8K followers
Sustained ~252–291 GiB/s of HBM memory throughput on H100s under decode load — essentially hitting the hardware roofline. This matters because HBM throughput, not FLOPs, is the real bottleneck in LLM inference. By keeping memory nearly fully saturated, I’ve unlocked far higher efficiency and throughput than standard engines. The result: 0.9–1.36M tokens/sec with ~0.1 ms first-token latency #AI #LLM #GPURouter #H100 #Inference #CostEfficiency #Innovation #DeepLearning #AIInfrastructure #HighPerformanceComputing NVIDIA AMD OpenAI Google Shilpa Kolhatkar Keith Strier a16z speedrun
5
-
Henry D. Wolfe
DaVega & Wolfe Industries… • 2K followers
Hexcel Adds Independent Director in Deal With Activist Investor "Hexcel added a new independent director to its board in a move that avoids a potential proxy fight for the aerospace materials company with activist investor Vision One Fund. "As part of an agreement between the two sides, Vision One will withdraw their nominees to the board that were to be voted on at the company's annual shareholder meeting. Vision One will also support Hexcel's board nominees, and agree to certain standstill restrictions and mutual non-disparagement provisions, according to a filing with the Securities and Exchange Commission on Wednesday. "The new director, former Kaman Chief Executive Officer Neal Keating, was appointed effective Tuesday." Keating's experience in aerospace may make him an ideal director choice. Yet from a bigger picture perspective, I wonder if this settlement is the right move for investors. More and more activist campaigns are settling rather than running the full process of a proxy fight with typically a greater number of directors nominated by the activist. Hopefully, activists are becoming like the public company boards they battle, i.e. more focused on reaching consensus than big results. #governancearbitrage #proxyfights #corporategovernance #activistinvesting #valuecreation https://lnkd.in/gazBUGFp
4
-
Jordan Steiner, CFA
Developer Capital • 3K followers
"Build the event you wish existed" That's what we at Monadical did last week at #NYTW. We wanted an AI Engineers discussion for Engineers. There's always a lot of events out there for VCs to network, or for startups to learn about G2M, but very little on lessons learned from actual engineers in the field. So that's the event we hosted. Big thank yous to our awesome panel, Roy Pereira, Ben Cohen and Corey J. Gallon. Here's the key takeaways and the AI tools we're using. 🚀 All three panelists independently called AI Agents the most transformative LLM application they’ve used. They specifically called out Claude 3.5 Sonnet for its accuracy and reliability. 🪨 We dug into how LLMs are “jagged”, not general. They can be shockingly good at some tasks and completely fail at others. Everyone agreed: good evaluations are critical (and hard.) 🧪 Corey noted how public benchmarks and reality are two different things. Most public evals are saturated or gamed. ♊ Ben emphasized that AI projects are actually two projects: building the tool and building the evaluation process. 🧱 We explored how falling dev costs may impact startup defensibility and labor demand. Roy shared that founders are already shifting strategies in response. ⚒️ In a world of daily AI launches, the panel discussed how they decide what’s worth attention, and what’s just noise. They called out tools like Goose, Aider, Claude Code, and Monadical’s own Cubbi, which helps run agentic workflows safely in dev environments. (links in the comments). CTA: What would you want to hear in an AI Applied Engineering talk you attended?
33
5 Comments -
Matt Rappaport
Berkeley Gateway Accelerator • 9K followers
Berkeley has long been a hub where deep physics intersects with real-world ambition, and photonics is exemplifying this in real time. Future Frontier Capital and the Berkeley Gateway Accelerator are co-hosting an evening focused on the future of light-based computing with Opticore. This event will bring together founders who are building the chips and researchers from the labs that trained them, creating a global ecosystem that connects North America to Europe's photonic infrastructure. Panelists include: - Ryan Hamerly, CTO of Opticore (Ph.D., Stanford; photonic chips at 100x GPU efficiency) - Daniel Klawson, CEO of TopoLight (Ph.D., UC Berkeley Integrated Photonics Lab; quantum computing photonics) - Jorn Smeets Managing Director NA of PhotonDelta (scaling the global photonic chip ecosystem) - Moderated by Kai Schuster Schuster of Opticore (materials scientist, photonics founder) The format features a concise panel discussion followed by a Q&A, with ample networking time to explore the topics that matter to you. If you are involved in hardware, tracking deep tech, or looking to understand the implications of photonics for AI infrastructure and quantum systems, this event is worth your evening. Thursday, May 7 | Doors at 5:30 PM | Berkeley Gateway Accelerator Link to register in the comments section.
67
3 Comments -
Matt Rappaport
Berkeley Gateway Accelerator • 9K followers
You can have groundbreaking science, elegant engineering, and even early customer interest. But without this one thing, you don't have an investable business. After two decades running IP strategy projects and now leading UC Berkeley's Deep Tech Innovation Lab while building the Berkeley Gateway Accelerator, I've watched countless brilliant technologies die in the valley between breakthrough and business. The problem isn't what most technical founders think it is. It's not about having better tech. It's not about getting more funding. It's not even about finding product-market fit. The fatal flaw shows up much earlier—and it's almost always the same mistake. In my latest piece, I answer questions from a Taiwanese entrepreneur about what really separates deep tech ventures that scale from those that stall. Including one provocative suggestion for Asian ecosystems that has nothing to do with technology. Read the full conversation, linked below. #DeepTech #Innovation #Startups #IntellectualProperty #VentureCapital
60
6 Comments -
Tom Carter
2K followers
My investment framework centers on 7 layers. CASIMIR fits into the first layer which is the Observation Layer. Out of 22 billion IoT devices on earth and ~10k satellites in space in 2026 we may have a 6 billion (and growing) ultra low power devices on earth and space that can one day run on MicroSparc 👍🏼 -> Eternal power chip Harold “Sonny” White + Capital Factory #HardThings #deeptech
4
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content