Eric H. Walker

Our Public Airtable Interface lets you search, sort, and filter through 8000+ #VC professionals. Discover essential details like firm names, websites, and social media links. https://lnkd.in/gyxeJQEm #startup #startups #venturecapital #funding #vc #privateequity

1

Please enjoy Rijul Gupta talk about election deep fake detection in front of the Senate Judiciary Committee highlighting the scale of the problem and the tech available to mitigate the risk. #deeptech #elections #deepfake https://lnkd.in/dXbH3jUC

3

Chart of the Day - Other Transactions (OTs) Spending by Emerging Tech Other Transaction Authorities are increasingly being used by agencies...but for which emerging tech categories and which agencies? This chart is the first of many on the Public Innovation Ecosystem Dashboard and it shows: - $48B in spending through OTs (FY18-FY23) - Trend chart coming - While Electronic Warfare and biotech are the largest categories, biotech is inflated due to 2020 and 2021 pandemic era spending. - Over the past two years, Electronic Warfare, Space Tech, Clean Energy, Advanced Comm and Data Science are up along with microelectronics and advanced materials. - DoD is by far the largest user, followed by HHS and Energy What else do you want to know? _____________________________________________________ About Chart of the Day and the Public Innovation Ecosystem Dashboard - Public Spend Forum will be releasing a chart with insights on the Innovation Ecosystem and the Industrial Base. You can access the Public Innovation Ecosystem Dashboard to see more information (link below). Link: https://lnkd.in/e4ekmXe4 You can also access more dashboards and details through PSF's Market Intelligence Office Subscription. #emergingtech #defensetech #Ota Jenna Roueche' Trish Martinelli Kevin Landtroop Greg Lewis Raphael Hernandez Brian Phillips

33

2 Comments

The Rise of SPVs in AI Startup Investments Special Purpose Vehicles (#SPVs) are increasingly popular among small investors, allowing them to pool their money to invest in coveted #startups like Anthropic, Groq, OpenAI, Perplexity, and #ElonMusk's x.AI. SPVs are formed by #investors with direct access to shares of these startups, who then sell a portion of their allocation to external investors, often charging significant fees and retaining a profit share. Early backers in sought-after #AI #startups exercise their pro-rata rights, creating opportunities for SPVs to acquire shares. VCs may offer access to SPVs to existing #limitedpartner investors or use brokers to reach a broader pool of potential investors. Some SPVs are formed on top of others, leading to complex layers of investment. Shares in certain startups, like Anthropic, became available due to events like FTX's bankruptcy auction, flooding the market with shares. SPVs can be formed alongside primary fundraising rounds, allowing small investors to participate simultaneously with major investors. While SPVs provide access to otherwise inaccessible investments, they come with high risks, as backers do not receive direct information about the companies. For more information https://lnkd.in/dqxCtget #AIInvesting #StartupFinance #SPVStrategy #VentureCapital #InvestmentTrends #TechInvesting #Entrepreneurship

4

Future Token Prediction Model FTP: A New AI Training Method for Transformers that Predicts Multiple Future Tokens The current design of causal language models, such as GPTs, is intrinsically burdened with the challenge of semantic coherence over longer stretches because of their one-token-ahead prediction design. This has enabled significant generative AI development but often leads to “topic drift” when longer sequences are produced since each token predicted depends only on the presence of mere preceding tokens, not from a broader perspective. This narrows the practical usefulness of these models in complex real-world applications with strict topic adherence, such as narrative generation, content creation, and coding tasks. Overcoming this challenge by enabling multi-token prediction would greatly improve semantic continuity, accuracy, and coherence of the generated sequences of the current generative language models. There have been various ways through which multi-token prediction has been addressed, each with different limitations. Models that aim to make predictions for multiple tokens by splitting embeddings or having multiple language heads are computationally intensive and often don’t perform well. For Seq2Seq models in encoder-decoder sets, while this allows for multi-token prediction, they fail to capture past contexts into one single embedding; hence, a lot of inefficiencies result. While BERT and other masked language models can predict multiple tokens of a sequence that are masked, they fail in left-to-right generation, hence restricting their use in sequential text prediction. ProphetNet, on the other hand, uses an n-gram prediction strategy; nonetheless, this is not flexible across a wide range of data types. The basic drawbacks of the aforementioned methods are scalability issues, computational waste, and generally unimpressive results while generating high-quality predictions over long-context problems. The researchers from EPFL introduce the Future Token Prediction model, representing a new architecture to create broader context-aware token embeddings. This will enable seamless multi-token predictions where, in contrast with standard models, the embedding from the top layers is used by a transformer encoder to provide “pseudo-sequences” cross-attended by a small transformer decoder for next-token predictions. In this way, the model leverages such encoder-decoder capability of the FTP for retaining context information from tokens of the previous history to make smoother transitions and maintain topic coherence across multi-token predictions. With more widespread sequence context encoded within its embeddings, FTP provides stronger continuity for generated sequences and has become one of the best approaches to content generation and other applications that require long-form semantic coherence. The FTP model employs a modified GPT-2 architecture that is made up of a 12-layer encoder with a 3-layer decoder. Its encoder...

1

Massachusetts' tech ecosystem thrives through a unique "DNA" of innovation—educational institutions, established companies, venture capitalists, and a deeply collaborative environment. With over 100 colleges and universities, including global leaders like MIT and Harvard, the region consistently ranks as one of the top innovation hubs in the world. An example of the strength of this DNA is in the biopharma industry. The 2023 Biopharma Funding & Pipeline Report stated that Massachusetts-based biopharma companies received $7.67 billion in venture capital funding, accounting for 31% of all national VC investments, second only to California. What can other states learn from what Massachusetts has created?

3

This should be a very interesting webinar discussing the intersection of AI and Blockchain.

4

When Leonard Tang took the stage during November's edition of #DataDrivenNYC (Matt Turck, FirstMark, Ramp HQ), you could feel the room shift. At 23, this guy isn’t just playing the game; he’s rewiring it. His confidence, sharp insights, and unapologetic drive made it clear: Haize Labs is here to shake things up. After the talk, I had the chance to meet him at their brand-new NYC headquarters, and what I saw blew me away. Developers, buckle up—Haize isn’t just a tool; it’s a damn revolution. Leonard doesn’t waste time. He’s tackling one of the gnarliest challenges in AI: trust. The kind of trust that can make or break your product, your brand, and your business. During his talk, he introduced Haize’s “haizing” suite—a next-gen dynamic testing tool designed to break your AI before the world does. But this isn’t just about finding bugs; it’s about building AI that performs when it counts, under pressure, and in the chaos of real-world deployment. This is where it gets exciting. Haize isn’t some shiny object in the AI space—it’s a workhorse. They’re solving the gritty, foundational problems that most companies ignore until it’s too late. OpenAI, Anthropic, AI21 Labs? They’re already clients. Enterprise players like Deloitte and MongoDB? Onboard too, as partners. Why? Because Leonard and his team are embedding trust into the DNA of AI systems, not as a bolt-on but as a baseline. The lesson here isn’t just for AI nerds; it’s for anyone scaling a business. Leonard’s message was crystal clear: trust isn’t something you earn later—it’s the foundation you build now. Whether it’s your product, your team, or your vision, reliability isn’t negotiable. Leaders, if you’re still playing defense, reacting to problems as they arise, you’re already losing. Invest in systems, tools, and people who think two steps ahead. That’s how you win. And developers, listen up. Haize is your cheat code. Their four-product suite—dynamic testing, monitoring, supervising, and robustification—isn’t just another add-on to your workflow. It’s a game-changer. Whether you’re tuning models or tightening up your prompts, Haize makes it faster, smarter, and frankly, more fun to build systems you can actually trust. Meeting Leonard and seeing Haize in action reminded me why companies like this matter. He’s not just building software; he’s building the future of AI. And here’s the kicker: he’s doing it with swagger, smarts, and a team that’s just as relentless as he is. #Networking #NYCTech #Data #DeveloperTools #AI #AIDriven #TechEcosystem #EnterpriseTech #EnterpriseAI #DeveloperPlatform #AISystems #OldDogLearnedNewTricks #Technology #Innovation Let’s connect and keep the momentum going across the tech ecosystem. Whether you’re a founder shaping the future, a leader driving change, a VC backing bold ideas, or an investor spotting the next big thing—together, we’re pushing boundaries. Proud to be building the future with you.

17

2 Comments

Had a great conversation with CEO.com Dove into some fun and heavy topics: *Current venture capital market *Top 2024 Marketing strategies *Why distribution is everything *Hyperlocal GTM strategy *The Importance of diversity in tech\ We also touched on the unique approach we take at Fiat Growth and Fiat Ventures to invest in, and navigate through this fast-changing landscape. Thanks to CEO.com for the opportunity to share my perspective.

34

Stepping into adjacent containers, digital, physical, magical to just, oh I don't know, keep it spicy. "Seira, a series of digital video artworks, is a synthesis of modern day technology and the yearning for profound human experiences. Utilizing TouchDesigner, AI software, code, and my original designs, I explore the intersection of the digital and the organic– drawing inspiration from technoromanticism, ancient Greek sirens, and synesthesia.   At its heart, this series is an exploration of the complex relationship between death and _l'appel du vide_—the call of the void. This psychological phenomenon, which tempts humans with the allure of self-destruction, is mirrored in the ancient myth of the sirens. The siren's song, like the void, is a lure that challenges us to confront our deepest impulses—the pull towards oblivion and the simultaneous yearning for transcendence. I intentionally draw a parallel between _l'appel du vide_ and the siren’s call, as both are manifestations of our intrinsic fascination with the unknown, with that which lies beyond the boundaries of safety and rationality." Seirá, Meditations on l'appel du vide (2024) by Neesh Chaudhary https://lnkd.in/gDm_3KTS

3

1 Comment

What do you get if you combine: marketing, education, and product feedback? You get Opus 2's ‘AI Labs’. AL takes a look at what Opus 2 is doing with its new AI Labs project for clients. Read on. #legaltech #legalAI #disputes https://lnkd.in/ei9w99nN

45

Congratulations to Abnormal Security on raising a $250M Series D financing at a $5.1B valuation, led by Wellington Management. We at Greylock are proud to be investing in Abnormal for the fifth time since Evan Reiser, Sanjay Jeyakumar and team started Abnormal in Greylock's offices. Abnormal's built the world's first AI-native cybersecurity platform that leverages AI to protect organizations most important vulnerability: humans and human behavior. Abnormal is also announcing today that they have crossed $200M ARR with 100% YoY revenue growth and is now used to secure nearly 20% of the Fortune 500. An important milestone as Abnormal advances its mission to protect humans using AI. CC: Asheem Chandna, Venky Ganesan, Stephen Ward, Rob Mazzoni https://lnkd.in/g4G4PRf9

176

4 Comments

Today, Destiny Tech100 Inc. filed a shelf registration statement authorizing a $1 billion capital raise upon effectiveness. This next step will accelerate our momentum in providing public access to private tech and continue building toward our target of 100 companies in the $DXYZ portfolio. We are excited to continue building this fund together, in public, but due to traditional rules and regulations, we cannot respond to inquiries at this time. Full registration statement: https://lnkd.in/gXB7SUrT The information in this prospectus (or Statement of Additional Information) is not complete and may be changed. The shares of our common stock sold in our capital raise can involve one or more transactions as needed to achieve our goal. We may not sell these securities until the registration statement filed with the Securities and Exchange Commission is effective. This prospectus (or Statement of Additional Information) is not an offer to sell these securities and is not soliciting an offer to buy these securities in any state where the offer or sale is not permitted. (disclosure: d.xyz/terms)

99

7 Comments

The business model and numbers of a corporate venture builder, Coplex - Venture Studios, launched 150+ startups (65+ are alive today) and builds studios for corporates: 1. Originally, it was a digital agency. Zach Ferres joined as CEO in 2012. In 2016, they transitioned into a venture studio, building startups with corporations (initially taking 5% equity in addition to fees). In 2020, they switched to the GP/LP funds model with corporates. 2. Build 10+* venture studios for corporates. 3. Each fund is $20-40M. 4. Now, the process starts with a $300-500K* charge for several months of studio design, resulting in a 120+ page document describing all processes of the future studio. 5. They always work with external ideas from corporates. 6. Equity split: on average, 20% studio, 30% corporation, 35% CEO, 15% ESOP. 7. The studio usually takes 20 to 30% carried interest in the fund. They don't charge a management fee but do charge an annual fee to run the studio. 8. When the studio co-invests alongside corporate partners, they can participate in the upsides as an LP. 9. On average, these funds invest $500K in each startup. The funds support building 5 to 10 companies a year for 3-4 years, with the remaining capital allocated for pro-rata follow-on investments. 10. They have a team of 8 full-time employees. Before COVID, they had 35 employees but decided to leverage vendors and freelancers instead of an in-house team. 11. They find the CEO for corporate studios and then hire a team. For startups, they prefer to attract non-technical and revenue-focused CEOs. 12. Ideal corporate partner: Minimum $1B in revenue; launched their CVC (a good sign); large and boring vertical businesses. Why partner with corporates? 1. They didn't want to have huge sales teams to attract entrepreneurs and investors. 2. Raising external capital is very difficult for studios. In 2016, it was so challenging to raise capital for a venture studio fund that they decided never to do it. This is where corporations come in with their capital. 3. As a studio, you don't need to charge startups. Challenges of working with corporations: 1. Competing with their internal functions. The studio must be outside the corporation because dependence on internal corporate staff makes it very hard to run a studio. 2. Long sales cycles to convince their boards to take the risk of building externally and then having a separate team. 3. Complex legal, tax, and structural implications, especially when working with finance, healthcare, insurance companies, and nonprofits. 4. It’s still a 10-year lifecycle. View the full interview for more insights: https://lnkd.in/dF-whjYR * Some updated numbers are from Michael Ellenby at last Friday's meeting of the Venture Studio Family. Join the paid community to get more studio insights and participate in regular online meetings with studios.

128

14 Comments

I recently joined New North Ventures as a General Partner. There are hundreds of Venture Capital firms investing in software, cloud, AI, etc. What makes New North Ventures different? DUAL USE - New North focuses on Defense/Intelligence technology startups with "dual use" crossover to Commercial customers. The Defense Department has published a list of Critical Technology Areas they are interested in. Examples of "dual use" companies in our Fund I portfolio include Hawkeye 360 and Hubble Network. You can see more examples on our Portfolio page.    VALUATION ARBITRAGE - Defense related startups tend to have low market valuations because their exit opportunities are sometimes limited to Defense Prime Contractors or related companies that don't command high stock market valuations. The arbitrage happens when these startups crossover to commercial markets where potential acquirers have market valuations that could be 10X higher than similarly situated defense companies. This is where the magic happens. EARLY STAGE - The Value creation curve is faster and easier at Seed stage or Series A. It is much easier to grow from $10M to $100M, and from $100M to $1 Billion, than it is later in the curve, to grow from $1B to $10B, and $10B to $100B. Investing at early stage is riskier, but the percentage returns can be higher. I have spent my entire professional career helping startups grow, raise money, and get to an exit. That experience can help improve the odds of success. NON DILUTIVE CAPITAL - The SBA just announced the SBICCT program in partnership with the DoD Office of Strategic Capital to provide 125% matching funds for investments we make in startups that support the Critical Technology Areas. It is a non-dilutive loan to our fund, with a low interest rate, that is paid back at a liquidity event, but the SBA doesn't participate in the capital gains of the transaction. This new program can double Limited Partner returns over other non-leveraged VC funds. PORTFOLIO BALANCE - Capital allocation balance is important to LP investors. Many VC funds have the same general thesis and invest in similar software tech companies. Those funds don't tend to have exposure to Defense/Intelligence startups with crossover dual use potential. DOUBLE BOTTOM LINE - At this stage in my career and life it isn't about the money. It is about the mission. Supporting our national security interests is my highest priority, and most important mission. It isn't often that you can invest your time and money to get a great financial return AND serve your country at the same time. This is a double bottom line I can get excited about. TEAM - The New North team is amazing. Our Partners have experience at Google, Microsoft, CIA, Palantir, Dyn, Minim, and Advisors from NASA, CIA, US Army, and National Security Advisors. Jeremy Hitchcock Stephanie Usry Bellistri Rob May Kevin Higgins Susan Still Kilrain Rick Waddell Marc Wolf Jessica Nelson My new email is [email protected]

151

16 Comments

This is a thought-provoking interview, Tomer. I have listened to the book by Deepak Chopra MD (official) called Digital Dharma, and it is fascinating to have a different perspective on the use of AI. Those of us who are immersed in helping businesses improve daily tasks, systems, content, research etc, often forget that there are other positive uses for individuals (of all ages). I have spent many hours opening the world of AI possibilities to seniors, and no matter how much I have learned, other uses always pop up. I am constantly amazed by where an open mind and genuine curiosity can lead in the ever-evolving AI world we are living in. #ArtificialIntelligence #AIForGood #DigitalInclusion #OpenMind #LinkedInCommunity

1

Tonya J. Long brings us insights and observations that are worth spending some of our precious weekend brain cells considering. We are at a #crossroads. #AI is changing everything, not just #technology products. The ability to #scale is dramatically improved if we take advantage of it. The NEED to scale leaves us with questions: ➡️ Scale bandwidth? ➡️ Scale customer base? ➡️ Scale talent? ➡️ Scale innovation? Tonya brings us the best and brightest moments from a reinvigorated silicon valley.

3

1 Comment

🗞️ Breaking News: Red Ventures Selling CNET at a huge Loss❗️ Context: Red Ventures, the privately held digital media and marketing juggernaut based in Fort Mill, South Carolina, is selling CNET, one of the largest tech publishers on the web, to Ziff Davis for just $100 million. For context: Red Ventures bought CNET for approximately $500 million. Reason for the Sale: Total traffic to CNET has dropped by a staggering 50% over the last three years. The main driver behind this decline is a 68% drop in organic traffic (loss off Google rankings and traffic). Tables turned: This sale is a stunning turn of events for CNET, which once bought Ziff Davis (then a tech magazine firm) in a deal worth roughly $1.6 billion - more than two decades ago. Catch Up: - Ziff Davis trades publicly with a market value of $1.94 billion. The company lost about half of its marketcap over the course of the last 48 months. Strategic rationale: If Ziff Davis is able to ink a content licencing deal with Google, Bing/Microsoft or ChatGPT (or else e.g. X.AI) - like Reddit did - it might turn out to be a very smart deal …

5

1 Comment