StarZero All shows Try the podcast clipper →

Dec 31, 2025 · 22m · latent-space

[State of Research Funding] Beyond NSF, Slingshots, Open Frontiers — Andy Konwinski, Laude Institute

Andy Konwinski · 17m spoken Shawn Wang · 2m spoken
0:00 / 0:00
▶ Watch on YouTube →

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

Databricks and Perplexity co-founder Andy Konwinski discusses the launch of the Laud Institute, detailing its mission to revitalize open-source AI research and empower academic founders through venture-speed philanthropic funding, compound AI systems innovation, and cross-institutional coalitions.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 13.3% of the talking time here. How this is scored →

The hosts as informed peer 4.4 Guest teaching 5.0 Guest disagreement 3.0 The hosts pushing back 3.4
05100:0010:0020:003:28–6:36 · The hosts as informed peer 4/10 The Academic Lab to Startup Pipeline and Founder Dynamics Swyx presses on whether Databricks' eight-cofounder model is replicable for typical Laude grantees. Andy challenges the premise by citing multiple precedents like OpenAI, VMware, and Snorkel to argue research teams naturally de-risk founder conflict.6:37–10:25 · The hosts as informed peer 5/10 Slingshot Research Directions: Compound AI Systems and JEPA Swyx demonstrates technical familiarity with prompt optimization and DSPy, prompting Andy to delineate DSPy as an agent compiler versus JEPA as an evolutionary genetic optimizer.10:26–13:32 · The hosts as informed peer 4/10 Expanding Academic Entrepreneurship Networks Beyond the West Coast Swyx challenges whether Laude is over-indexed on Berkeley and Stanford. Andy rejects the geographic limitation, listing their graduate clubs and grantee pipelines across UIUC, CMU, UW, Wisconsin, and Canadian labs.13:32–17:01 · The hosts as informed peer 4/10 Reimagining Research Funding: Comparing Laud to the NSF Model Swyx asks what is broken about the NSF funding process. Andy forcefully rejects the framing, praising NSF's historical return while detailing why computer science funding caps require Silicon Valley-style high-velocity grantmaking.17:02–22:21 · The hosts as informed peer 5/10 Open Frontiers Initiative and the Future of Open AI Research The conversation shifts to a collaborative discussion on the Open Frontiers initiative and Western labs falling behind Chinese startups in open-source research publication.3:28–6:36 · Guest teaching 5/10 The Academic Lab to Startup Pipeline and Founder Dynamics Swyx presses on whether Databricks' eight-cofounder model is replicable for typical Laude grantees. Andy challenges the premise by citing multiple precedents like OpenAI, VMware, and Snorkel to argue research teams naturally de-risk founder conflict.6:37–10:25 · Guest teaching 5/10 Slingshot Research Directions: Compound AI Systems and JEPA Swyx demonstrates technical familiarity with prompt optimization and DSPy, prompting Andy to delineate DSPy as an agent compiler versus JEPA as an evolutionary genetic optimizer.10:26–13:32 · Guest teaching 5/10 Expanding Academic Entrepreneurship Networks Beyond the West Coast Swyx challenges whether Laude is over-indexed on Berkeley and Stanford. Andy rejects the geographic limitation, listing their graduate clubs and grantee pipelines across UIUC, CMU, UW, Wisconsin, and Canadian labs.13:32–17:01 · Guest teaching 6/10 Reimagining Research Funding: Comparing Laud to the NSF Model Swyx asks what is broken about the NSF funding process. Andy forcefully rejects the framing, praising NSF's historical return while detailing why computer science funding caps require Silicon Valley-style high-velocity grantmaking.17:02–22:21 · Guest teaching 4/10 Open Frontiers Initiative and the Future of Open AI Research The conversation shifts to a collaborative discussion on the Open Frontiers initiative and Western labs falling behind Chinese startups in open-source research publication.3:28–6:36 · Guest disagreement 4/10 The Academic Lab to Startup Pipeline and Founder Dynamics Swyx presses on whether Databricks' eight-cofounder model is replicable for typical Laude grantees. Andy challenges the premise by citing multiple precedents like OpenAI, VMware, and Snorkel to argue research teams naturally de-risk founder conflict.6:37–10:25 · Guest disagreement 2/10 Slingshot Research Directions: Compound AI Systems and JEPA Swyx demonstrates technical familiarity with prompt optimization and DSPy, prompting Andy to delineate DSPy as an agent compiler versus JEPA as an evolutionary genetic optimizer.10:26–13:32 · Guest disagreement 3/10 Expanding Academic Entrepreneurship Networks Beyond the West Coast Swyx challenges whether Laude is over-indexed on Berkeley and Stanford. Andy rejects the geographic limitation, listing their graduate clubs and grantee pipelines across UIUC, CMU, UW, Wisconsin, and Canadian labs.13:32–17:01 · Guest disagreement 5/10 Reimagining Research Funding: Comparing Laud to the NSF Model Swyx asks what is broken about the NSF funding process. Andy forcefully rejects the framing, praising NSF's historical return while detailing why computer science funding caps require Silicon Valley-style high-velocity grantmaking.17:02–22:21 · Guest disagreement 1/10 Open Frontiers Initiative and the Future of Open AI Research The conversation shifts to a collaborative discussion on the Open Frontiers initiative and Western labs falling behind Chinese startups in open-source research publication.3:28–6:36 · The hosts pushing back 4/10 The Academic Lab to Startup Pipeline and Founder Dynamics Swyx presses on whether Databricks' eight-cofounder model is replicable for typical Laude grantees. Andy challenges the premise by citing multiple precedents like OpenAI, VMware, and Snorkel to argue research teams naturally de-risk founder conflict.6:37–10:25 · The hosts pushing back 3/10 Slingshot Research Directions: Compound AI Systems and JEPA Swyx demonstrates technical familiarity with prompt optimization and DSPy, prompting Andy to delineate DSPy as an agent compiler versus JEPA as an evolutionary genetic optimizer.10:26–13:32 · The hosts pushing back 5/10 Expanding Academic Entrepreneurship Networks Beyond the West Coast Swyx challenges whether Laude is over-indexed on Berkeley and Stanford. Andy rejects the geographic limitation, listing their graduate clubs and grantee pipelines across UIUC, CMU, UW, Wisconsin, and Canadian labs.13:32–17:01 · The hosts pushing back 4/10 Reimagining Research Funding: Comparing Laud to the NSF Model Swyx asks what is broken about the NSF funding process. Andy forcefully rejects the framing, praising NSF's historical return while detailing why computer science funding caps require Silicon Valley-style high-velocity grantmaking.17:02–22:21 · The hosts pushing back 1/10 Open Frontiers Initiative and the Future of Open AI Research The conversation shifts to a collaborative discussion on the Open Frontiers initiative and Western labs falling behind Chinese startups in open-source research publication.

speaking balance: gold is the hosts, purple is the guest (3 minute bins)

0:00 · the hosts 15.8% · guest 84.2%0:00 · the hosts 15.8% · guest 84.2%3:00 · the hosts 17.7% · guest 82.3%3:00 · the hosts 17.7% · guest 82.3%6:00 · the hosts 14.3% · guest 85.7%6:00 · the hosts 14.3% · guest 85.7%9:00 · the hosts 12.9% · guest 87.1%9:00 · the hosts 12.9% · guest 87.1%12:00 · the hosts 13.1% · guest 86.9%12:00 · the hosts 13.1% · guest 86.9%15:00 · the hosts 16.2% · guest 83.8%15:00 · the hosts 16.2% · guest 83.8%18:00 · the hosts 3.2% · guest 96.8%18:00 · the hosts 3.2% · guest 96.8%21:00 · the hosts 14.2% · guest 85.8%21:00 · the hosts 14.2% · guest 85.8%
Sharpest disagreement ▶ 13:38 Premise rejection on NSF dysfunction

Andy directly shuts down Swyx's framing of a broken NSF, clarifying that NSF is underfunded rather than conceptually flawed.

 Hardest push from the hosts ▶ 10:36 Pushback on geographic insularity

Swyx presses Andy on whether Laude's deal flow is disproportionately restricted to West Coast elite institutions.

 Biggest teaching moment ▶ 13:45 Breakdown of NSF CS funding constraints

Andy details exact NSF budgetary constraints ($1B dedicated to CS out of $10B) and explains how venture picker dynamics solve capital deployment bottlenecks.

 The host holds their own ▶ 8:39 Host technical framing on DSPy architecture

Swyx sharpens the technical discussion by distinguishing DSPy's framework branding from JEPA's optimizer role.
the scores for every segment, with the reasoning behind each
ChapterTopicThe hosts as informed peerGuest teachingGuest disagreementThe hosts pushing backWhy
The Academic Lab to Startup Pipeline and Founder Dynamics 4544 Swyx presses on whether Databricks' eight-cofounder model is replicable for typical Laude grantees. Andy challenges the premise by citing multiple precedents like OpenAI, VMware, and Snorkel to argue research teams naturally de-risk founder conflict.
Slingshot Research Directions: Compound AI Systems and JEPA 5523 Swyx demonstrates technical familiarity with prompt optimization and DSPy, prompting Andy to delineate DSPy as an agent compiler versus JEPA as an evolutionary genetic optimizer.
Expanding Academic Entrepreneurship Networks Beyond the West Coast 4535 Swyx challenges whether Laude is over-indexed on Berkeley and Stanford. Andy rejects the geographic limitation, listing their graduate clubs and grantee pipelines across UIUC, CMU, UW, Wisconsin, and Canadian labs.
Reimagining Research Funding: Comparing Laud to the NSF Model 4654 Swyx asks what is broken about the NSF funding process. Andy forcefully rejects the framing, praising NSF's historical return while detailing why computer science funding caps require Silicon Valley-style high-velocity grantmaking.
Open Frontiers Initiative and the Future of Open AI Research 5411 The conversation shifts to a collaborative discussion on the Open Frontiers initiative and Western labs falling behind Chinese startups in open-source research publication.

Statements from this episode (8)

Assertion Not checkable as stated
Jeff Dean and Databricks founders invested in Konwinski's venture fund
“So the venture funding, there's 50 professors and PhDs, the top names like Jeff Dean and the top faculty at Berkeley and Stanford and my co-founders of Databricks and Perplexity who've invested in the fund.”
Andy Konwinski Dec 31, 2025 ▶ 2:43
Insight
Academic lab research teams de-risk founder divorce for venture capitalists
“Years of deep scars from working together, knowing each other's ins and outs and team, like the dynamism and like the gestalt, like some is greater than the parts of working together as a team. They're great bets. So you kind of de-risked the founder divorce r…”
Andy Konwinski Dec 31, 2025 ▶ 4:23
Insight
LLMs are commoditizing like raw compute, shifting value to abstraction layers
“Language models themselves are more like compute or GPU a generation ago, where what can we build at the layer above? And in software systems, we've traditionally thought of VMware being a great example. You have the operating system and the underlying archite…”
Andy Konwinski Dec 31, 2025 ▶ 7:04
Opinion
The National Science Foundation has returned thousands of times its public cost
“It has been probably the best investment the American popular public has ever made. Thousands X return on their investments. I mentioned Google earlier, Databricks, all these researchers came from NSF funding”
Andy Konwinski Dec 31, 2025 ▶ 13:42
Assertion Not checkable as stated
Frontier AI research requires up to $100B, far exceeding NSF budgets
“It was one billion dollars a year for computer science and that they're trying to cut that into half of that, but we need 10 to a hundred billion dollars to do frontier AI research.”
Andy Konwinski Dec 31, 2025 ▶ 14:12
Opinion
Western open AI research has lost the top spot to China
“Western open science and research discourse has no, has lost the number one spot to China.”
Andy Konwinski Dec 31, 2025 ▶ 18:56
Assertion Contradicted
All major US AI labs stopped publishing research after OpenAI closed
“Whereas in the United States, since OpenAI closed their doors and stopped publishing, so did all the other labs.”
Andy Konwinski Dec 31, 2025 ▶ 19:25
Disclosure
Open Frontiers will gather the top 100 open AI researchers in SF
“The idea is like, let's team up. Get together one day in San Francisco in the next five months and get the hundred most influential open researchers together for conference and for meetings where we share our roadmaps and we talk about common goals amongst the…”
Andy Konwinski Dec 31, 2025 ▶ 20:43
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 200 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more