Jacob G.

At a time of widespread concerns about the safety, efficacy and economics of AI, venture capital pioneer Vinod Khosla is still all in. Yes, artificial intelligence will become capable of replacing a good deal of human labor, Khosla said in my column for this week. But that will push down the price of healthcare, education and other services. Many current forms of work will even be eliminated, but society will be able to create a more robust safety net than is possible today. “Most expertise in the world, whether you’re talking about structural engineers, oncologists, mental health therapists or primary care doctors, or journalists and teachers, that expertise will be near-free for all of us to access,” Khosla said during a conversation at Khosla Ventures in Menlo Park. About 80% of the work involved in 80% of jobs across the economy can be automated over time, according to Khosla. “So 64% of all jobs can be done by an AI,” he said. Over a 25- or a 50-year period, GDP growth has the potential to climb to 5%. During the second quarter of this year, by comparison, the economy grew at an annualized rate of 2.8%. “We will have in that excess GDP growth lots of room to take care of the people who will be adversely affected economically,” he said. “If medical services are a lot cheaper, if education services are near free, if eldercare becomes substantially cheaper—and today eldercare is a liability that’s looming for so many nations around the world at a level that people haven’t accounted for—the social safety net is much easier to construct,” said Khosla. Khosla isn’t alone in his outlook, of course—at least elements of it. Existing generative AI and other technologies have the potential to automate work that absorbs 60% to 70% of employees’ time today, consulting giant McKinsey & Company said in a June 2023 report, raising its estimate from roughly half of employees’ time in a pre-generative AI study. That potential won’t necessarily be fully realized anytime soon, though, given the effort it takes to integrate new technology and the need for that work to make economic sense and pass regulatory scrutiny. Perhaps 30% of today’s work hours might be automated by 2030, according to McKinsey Global Institute Partner Michael Chui, an author of the report. 

30

3 Comments

In this week's column I look at why AI can't be adopted in isolation -- it must occur in a technology environment that can manage the unprecedented demand for data. The rise of AI is analogous to an electric car, which requires a total reconsideration of many components in a gasoline-powered car, right down to the tires. To gain traction, AI will need a new set of wheels, too. “Our vision was to build infrastructure for these new AI workloads,” said Chief Executive Renen Hallak, who founded VAST DATA in Israel in 2016. In December, VAST said it secured $118 million in a Series E funding round, led by Fidelity Management & Research, that nearly tripled its valuation to $9.1 billion. Data storage historically has been organized in tiers, in which recent, high-priority data was kept readily accessible and older data was buried further down, according to Hallak. “That’s not the case anymore with these new AI workloads,” Hallak told me in an interview at the company’s office in New York. “Once you have a good AI model you want to infer on all of your history, because that’s how you get value out of it. And then as you get more information you want to retrain and build a better model,” Hallak said.  “You read data over and over and over again across many petabytes and in some cases exabytes. And so that’s a very different problem." This shift is already under way at some highly data-intensive companies such as Pixar Animation Studios, the Disney movie studio behind this summer’s hit “Inside Out 2.” It has been working with VAST since 2018. Beginning with its 2020 film “Soul,” Pixar has employed a technique known as volumetric animation that produces more highly detailed characters and movements. The use of volumetric animation was more extensive in its 2023 release “Elemental,” for which Pixar used AI to curate protagonist Ember Lumen’s flames. “Inside Out 2,” which came out in June, had double the data capacity needs of “Soul” and required about 75% more computation. Pixar’s old system of moving data from high-performance drives to lower-performance drives when not in use didn’t work for rendering volumetric characters, according to Eric Bermender, head of data infrastructure and platforms at Pixar. For AI, Pixar tends to employ on-premises networks as opposed to the cloud. AI in general doesn’t easily conform to traditional architectures, Bermender said. “These workflows tend to require processing massive amounts of diverse data that is not cacheable, not sequential, and which traditionally would have been stored in lower performance archive tiers,” he said. Do you think your company’s information systems are keeping pace with the increasing demands of AI and data? Share your thoughts.

55

2 Comments

It's great to see Victor Dibia, PhD and Chi Wang started writing a book on AutoGen with Manning Publications Co.! Building multi-agent systems is a frontier topic now, and AutoGen is one of the most interesting frameworks in the field. https://mng.bz/mRJM I had some experience writing multi-agent systems with AutoGen, and I like it. I truly believe in agentic workflows, as collaborative agents are the next step after using plain LLMs (even with all the magic of Chain-of-Thought, Self-Consistency, RAG, and even brand-new Reflection and Reasoning). Do not undervalue materials from the authors of AutoGen and AugoGen Studio! They know what they are talking about!

52

5 Comments

A couple of weeks ago a journalist at our national broadcaster did a long piece on driverless taxis in San Francisco. While the piece played up on peoples' fears and asked if you'd feel safe in one, they never provided any safety data comparing driverless vehicles to those driven by people. I managed to reach them on Twitter and here's is what they said: "I agree that data is important. Unfortunately, there are no independent studies available. The only data is provided by robotaxis operators. One example from Waymo here (below)." I can't say how accurate these Waymo numbers are but they certainly sound impressive. I'm left with two questions. 1) If the driverless car operators' number are this good why don't they provide complete open access to their live numbers with independent researchers/journalists? 2) I'd be curious to know the number of accidents that are seen for every mile or hour of driving that have been recorded for human driven vehicles. And specifically, how do these numbers compare with those of the operators' claims. Insurance companies must have this information and feel more or less confident on insuring driverless car companies vs. people. At some point, if the driverless cars are truly that much better this should become a simple decision. https://lnkd.in/g6g3PqCa

6

8 Comments

I find it hard to believe that with millions of EVs entering the market and the proliferation of AI compute that PGE is actually going to have excess capacity in 2040 without any additional generation from 3rd parties. I am also curious if there are enough transmission lines and if PGE views that as someone else’s problem. Maybe they are planning on AI and ACTUAL figuring it all out! https://lnkd.in/gybsMmaw

10

5 Comments

Friends, Q2 was a quarter of progress and learning for us at HawkTower, where we invest in early-stage startups applying breakthrough computing technologies to maximize California's natural resource markets. Our focus is on technology-enabled interventions (e.g., AI-enabled geospatial radars for lithium-ion batteries in the waste stream) for critical, market-level imperatives (e.g., disaster prevention) that can drive rapid commercial adoption (e.g., placed on every recycling truck in the world). Our current investment priorities are: - Agricultural productivity and efficiency - Biomanufacturing - Labor and mobility automation - Waste stream optimization and decarbonization - Disaster prevention and management Here’s a snapshot of what we’ve been up to: Milestones and Investments: - In late April, we officially launched our fund, filing with the SEC. Yay. - By June, we completed our first fund close ahead of schedule, with backing from family offices, foundations, high-net-worth individuals, and organizations aligned with California's future. - Our portfolio now includes three early-stage investments (Farm-ng, Agriful Software, and Nexstera Tech). Strategic Partnerships: - We believe that startups operating closer to their customers and resources, especially in agriculture and biomanufacturing, have a strategic advantage. For instance, the Salinas Valley, a major agricultural hub, is less than an hour from Silicon Valley and serves as a center of thought leadership in the convergence of agriculture and AI (from Western Growers, Taylor Farms, and others). - Similarly, the North San Joaquin Valley is emerging as a biotech and agtech innovation hub, and we are proud to support BEAM Circular, a startup hub for the circular bioeconomy. Key Relationships: - Karen Warner, CEO of BEAM Circular, joined HawkTower as a strategic advisor. - Jackie Cruz, Ed. D. Cruz, Hartnell College’s VP of Advancement, advises on Salinas Valley and workforce development. - Ben Palone from Western Growers, with his technical background in agriculture, became a trusted advisor. - Matthew Hoffman from Driscoll's Global R&D team provides insights on global agriculture, product positioning, and investment strategy. - We collaborated with Monterey Bay DART for workforce development and automation in agtech. - Dominic Milano and the Milano Technical Group from Merced collaborated on key initiatives and advised us on investments. - HawkTower is participating in Santa Cruz Works's accelerator cohort as a mentor and advisor. Thank you Doug Erickson. - We published an overview of entrepreneurship across the UC system with input from across the University of California system. Moving into the fall, we remain deeply committed to supporting founders shaping California's future natural resource markets. Stay tuned and be in touch! : ) Db

95

5 Comments

Breaking down the Seattle Venture Ecosystem! Kudos to Jacob Laes and Nick Khormaei for shedding light on the often complex and opaque Seattle funding ecosystem with their comprehensive investor map. For founders, navigating the venture capital world can feel like a maze, but this map offers a much-needed clearer view of where Seattle’s investors are truly placing their bets. And you know me—data is everything! Data drives decisions and, in this case, holds the key to unlocking new opportunities for startups. This project covers 58 firms with insights on: Industry sector focus Investment stages Portfolio track records It’s a game-changer for founders looking to connect with the right investors more efficiently. Check out their breakdown for a detailed view, and feel free to reach out for the full data set! Evan Poncelet Jesse Posey #Startups #Entrepreneurship #VentureCapital #SeattleTech #Funding #TechInnovation #Investing #StartupCommunity #VentureEcosystem #FounderJourney #BusinessGrowth #SeattleStartups

10

1 Comment

"United Independents, the group behind the Independent National Convention set for Denver in September, has launched a new software platform for indie candidates called Electify.us, which came out of a partnership with incubator Wolfram Labs — the first tool to emerge from its program." I am so proud to share that Chief Product Officer Dana Cichon has launched the first political tech platform to come out of our accelerator program. What are you waiting for? We have the formula for breaking into the digital campaign space. Are you are the next success story? Read more in Campaigns & Elections -->> https://lnkd.in/eHhqa-Xr #techaccelerator #startups #incubator #VC #SaaS #software #political #campaigns

17

To all #climatetech #founders and #funders in NYC and beyond, please join us on June 5th for an evening of #breakthrough #decarbonization #innovation at the ReGen Accelerator #DemoDay! 🎉 Get ready to be inspired by our cohort of #earlystage startups: ​ ▶ aich2 is engineering a novel #hydrogen ecosystem, built around Solid Hydrogen carriers. ​​ ▶ carbogenics produces sustainable #carbon adsorbents from difficult-to-recycle organic waste, including coffee cups and wastewater screenings ​​ ▶ TeamStack.ai is driving better performance by harnessing #AI, data, & the collective intelligence of teams. ​ ▶ Tesleon LLC uses compelling characters and #storytelling to instill compassion for communities and our planet in the next generation. ​ ▶ Tolam Earth is a trusted #marketplace for high-quality, high-impact carbon offsets. Don't miss out! Secure your spot below. Big thank you to Brown Rudnick LLP for hosting this event, our partners at ClimateHaven for their hard work and dedication during the entire program, and to our incredible network of #mentors and #facilitators, each of whom have dedicated time to working with these incredible founders this spring, including Jason Ambrose, Marina Rakhlin, Lauren Zujkowski, Stuart DeCew, Matt Puccini Andy Rooks, Nina Raab, and Bill Stephney. And of course Shai Tamary, Lars Korby and Sarah Cleveland, without whom we would not be able run such a #program. #innovation #sustainability #climateaction #accelerator

47

1 Comment

Ethernet always wins. It has already adopted most of the features of Infiniband. Now, Meta reports Ethernet is more reliable.

14

🇺🇸 The work done by the DemocracyNext and MIT Center for Constructive Communication teams (and so many more involved) is so very important. Citizens’ Assemblies and broader deliberative dialogue solutions IRL are critical to solving for and strengthening trust in one another (across any number of domains, organizations, and verticals). At the same time, technology advancements & related solutions are vital to augment, scale, and enable better problem solving. Examples (and by no means comprehensive) on the technology side include: ✅ AI / LLM briefings on complex topics for educational aspects to constituents / assembly members, etc ✅ Predictive “Sim City 2.0” before / after modeling engine of particularly policy proposals across complex systems to enable assembly members & policy makers to try to understand and visualize the before / after outcomes, etc. ✅ Automated policy writer and real-time policy checker in terms of actual outcomes / improvements ✅ Integrations into Github / Wikipedia as to all conversations, information, etc (used for: training of AI systems/solutions, open knowledgebase on solutions to issues that multiple constituents / communities / orgs may be facing, etc) ✅ Visual solutions / storytelling systems of before / after from human voices ✅ Anti-VUCA metric analysis and optimizations for success 🔖 “In a landmark achievement, MIT CCC and DemocracyNext have collaborated on the first-of-its-kind, tech-enhanced citizens’ assembly in Deschutes County, Oregon. 12,750 invitation letters were mailed to residents across Central Oregon. Of the respondents, 30 delegates were chosen via sortition (lottery), to be broadly representative of the area. Each of these delegates attended two weekends full of learning, deliberation, and voting. Their task? Using their lived experience and newly acquired knowledge to come up with recommendations which would address the intransigent issue of youth homelessness in their area. For the first time, these delegates agreed to audio record their small-group conversations in order to allow researchers to better understand their deliberation process. As an early glimpse into the tech enhancements, delegates were presented with AI-generated summaries of what they discussed two weeks prior. Moreover, they were able to see what other groups discussed, a level of transparency not previously possible when facilitators would take imperfect, handwritten notes.” https://lnkd.in/gWjx-abm Claudia Lucy Robbie Panthea Deb Dimitra Josh Kevin Brooke Helen Andrew Alan Michelle Akshay Matthew Norbert Frederico Amar Stuart Katie Glenn Russell Max Daniel Peter Burton Sue Ellen David Mark Celine Christina Nicholas Kara Karen Amanda Brett Kristin

54

6 Comments

Startup radar time! We’re back with another spotlight on startups making moves in Seattle: Govstream.ai, led by Safouen (Saf) Rabah, is using artificial intelligence to help government leaders improve the permitting and urban development process. Integrate, led by John Conafay, aims to help companies streamline program management for hardware development. Sound Games, led by Mike Schmid, is developing video games and “other things related to making games." Tamnoon, led by Marina S., offers a cloud security software platform boosted by AI. Yara AI, led by Joe Braidwood, uses advanced language models and deep memory features to help customers boost mental wellness. Read more about the startups and their leaders here: https://lnkd.in/gVi3KpgK

189

10 Comments

Misguided AI "safety" regulation will hamper innovation and ultimately our ability to protect against misuses by bad actors. This is a great episode by a16z Anjney Midha ⤵ Current proposed bill SB-1047 can cause Civil AND Criminal charges for some foundation model developers even if, not their models, but derived models on top of it cause malicious outcomes. This is like 𝐜𝐫𝐢𝐦𝐢𝐧𝐚𝐥𝐥𝐲 𝐜𝐡𝐚𝐫𝐠𝐢𝐧𝐠 𝐭𝐡𝐞 𝐢𝐧𝐯𝐞𝐧𝐭𝐨𝐫 𝐨𝐟 𝐏𝐲𝐭𝐡𝐨𝐧 𝐢𝐟 𝐚 𝐩𝐢𝐞𝐜𝐞 𝐨𝐟 𝐏𝐲𝐭𝐡𝐨𝐧 𝐜𝐨𝐝𝐞 𝐢𝐬 𝐮𝐬𝐞𝐝 𝐦𝐚𝐥𝐢𝐜𝐢𝐨𝐮𝐬𝐥𝐲! The more appropriate course of action should be: - 𝐟𝐨𝐜𝐮𝐬 𝐨𝐧 𝐚𝐧𝐝 𝐭𝐚𝐫𝐠𝐞𝐭 𝐦𝐚𝐥𝐢𝐜𝐢𝐨𝐮𝐬 𝐮𝐬𝐞-𝐜𝐚𝐬𝐞𝐬 𝐚𝐧𝐝 𝐛𝐚𝐝 𝐚𝐜𝐭𝐨𝐫𝐬 rather than focusing on foundational models. - 𝐟𝐨𝐜𝐮𝐬 𝐨𝐧 𝐦𝐚𝐤𝐢𝐧𝐠 𝐫𝐨𝐛𝐨𝐭 𝐨𝐮𝐫 𝐬𝐞𝐜𝐮𝐫𝐢𝐭𝐲 𝐬𝐲𝐬𝐭𝐞𝐦𝐬 which are increasingly under attack, rather than being distracted by science-fiction and theoretical safety conversations. - 𝐚𝐜𝐜𝐞𝐥𝐞𝐫𝐚𝐭𝐞 𝐨𝐩𝐞𝐧 𝐬𝐨𝐮𝐫𝐜𝐞 𝐜𝐨𝐥𝐥𝐚𝐛𝐨𝐫𝐚𝐭𝐢𝐨𝐧 to foster a collaborative and open development. “We’ve barely just invented the printing press” let’s not stifle it by threatening to criminally charge the platform developers for what a very few bad actors will do with the new technology. Read the bill SB-1047. If you’re in California, please act so this bill as it reads currently doesn’t pass and get adjusted to address bad actors not the whole industry, because you set the stage for the rest of us!

8

Linxin Gu great article. thanks for sharing. Pls let us know folks in Boston area can help with your journey

4

3 Comments

Exowatt launches today! ☀️ The power requirements to scale AI are astronomical. Within a decade AI will require 650TWh of new power generation – more than the electricity demand of South Korea. In the U.S. alone, data centers will require the equivalent amount of electricity to power 40 million households by 2030. Exowatt P3s can generate electricity at scale at just a $0.04 LCOE and with entirely renewable tech. That’s lower than the cost of most fossil fuels and doesn’t account for subsidies. We have already amassed a 1.2 GW backlog, worth $Bs from the mega tech hyperscalers, energy and data center companies. The system works by capturing the sun’s energy in the form of heat, storing it in a heat battery and converting it to electricity via a heat engine even when the sun is no longer out. This changes the economics of solar by solving a problem called the “Duck Curve” where all conventional solar producers are forced to sell their power at the same time while at maximum output, driving the electricity cost sometimes to zero or negative intraday and leaving no supply at night. Data centers need to operate 24/7 and typically have 60% of their OPEX from energy costs, so an inexpensive and ever available alternative is very important to them. Data centers are straining to bring capacity online and without solutions like Exowatt, are being forced to turn to fossil fuels like coal, oil and natural gas. We hope you’ll join us in pushing forward an abundant and non fossil fuel powered future!⚡️ https://lnkd.in/ewcUpW9t

389

23 Comments

Earlier this week Texture, led by CEO & co-founder Sanjiv Sanghavi, announced its $7.5m seed round of fundraising with participation from Abstract Ventures, Day One Ventures, Equal Ventures, Lerer Hippeau, and a handful of angels. Dorothy Chang and I feel fortunate to count ourselves among the “handful of angels.” We didn’t invest just because Sanjiv is a close friend - though knowing him for over a decade has given us a lot of insight into his mind and how he operates. We invested for a lot of the typical reasons a VC would: Solving big problems with big potential returns - Venture investments only make sense if they have the chance to make massive returns. Texture is building the data infrastructure for energy networks - becoming a common data collection and sharing platform for every player in the energy grid, no matter the type of energy. There is already a huge gaping problem and a huge market here. Building for an inevitable future - On a macro level, as a society, we are moving more and more toward complex, interconnected energy sources as we optimize for the right balance of efficiency, sustainability and costs. Unique insight and founder-market fit - At a first glance, Texture seems like a big departure from what we originally knew Sanjiv from, as a Co-Founder of ClassPass. But since then, he has also played roles as the Chief Product officer at Arcadia, a successful energy data company (Arcadia’s CEO is also an investor in Texture) as well as an investor at Day One Ventures, investing in the energy/climate tech sector.  More details in this week's post: https://lnkd.in/eGBzXHwC

44

4 Comments

Top Tech Trends of 2025 Andreessen Horowitz (a16z) has spotlighted transformative tech ideas set to shape industries in 2025: 📈 AI-Powered Energy Needs: Reviving nuclear energy for resilient grids. 🖥️ Hardware-Software Synergy: Engineers blending AI with manufacturing. 🚀 Space Innovation: Reusable rockets enabling faster transport and new industries. 🤖 Decentralized Defense: AI-powered drones shaping military tech. 🌐 XR for Developers: AR/VR tools like Vision Pro driving spatial computing. 🌍 Earth Observation Data: Satellites unlocking new industry solutions. 🤝 Generalized Robots: Robotic data collection to accelerate AI. 📡 Optical Communications: Faster, more direct data transmission. Explore the full insights here: https://lnkd.in/gMT4nbhz #TechTrends #AI #Innovation #FutureTech #CFOInsights

4

Very excited for the Model Context Protocol, an open standard developed by Anthropic to connect AI systems to data sources. Open is the way to go! https://lnkd.in/eTwSXRfd

85

3 Comments