Danny Bernstein left Big Tech to fund farm-bots | E2310

Summary

Danny Bernstein spent close to 20 years in Silicon Valley — ten of them at Google — before leaving big tech to study “complex outdoor industries” and build Reservoir, the first on-farm robotics incubator. Joining live at dawn from a 40-acre campus in Salinas, California, he lays out the crisis driving the whole venture: over 400,000 farm jobs were posted last year and fewer than 1% got a single domestic applicant, while labor accounts for 40% of specialty-crop production costs — up to 60-70% for strawberries and 80% for table grapes. Reservoir runs as two interrelated entities: Reservoir Farms, a two-year-old operating company doing $5-7M ARR by renting lab-style farm access to 20 startups at $3,000-$6,000/month, and Reservoir VC, a $3M proof-of-concept fund. Bernstein frames domestic ag-tech as national security, pointing to how US-subsidized South American agricultural infrastructure (the “Blueberries Peru” story) left America importing more fresh produce than ever.

After Bernstein’s segment, the show pivots to a running feud over AI and design. Jason recaps Gal Shir’s viral “AI beat me at design” tweet, his own unclear reply that Figma CEO Dylan Field misread, and the “J-Trade” that resulted — Jason bought $100,000 of Figma in the secondary market. Bernstein (who stays on as co-host) makes the bull case for Figma: a driven founder, a loyal designer base, and a $10B currency to buy teams and move into Adobe’s adjacencies. Both agree AI raises the “average” design from roughly a 6 to an 8.5, making elite human designers who stand out from “AI slop” more valuable, not less.

The back half tackles the Department of Education’s new earnings test tying federal loan eligibility to graduate outcomes (Bernstein argues it doesn’t go far enough and colleges should be on the hook for defaults), the Paul Graham-shared “cheating chart” showing students who aced remote midterms then cratered on in-person finals, and Jason’s fury at “lazy” teachers who let it happen. It closes on an Off Duty run of Taylor Sheridan shows, the mortality memoir Not Fade Away, and Jason’s new 60-inch snake grabber for the ranch.

Highlights

”Less than 1% received a domestic applicant”

“These are jobs that domestic workers do not want to do. That’s categorical and clear. The American Farm Bureau said that there were over 400,000 farm jobs that were posted last year and less than 1% of them received a domestic applicant.” — Danny Bernstein, 6:00

yt-dlp --download-sections "*6:00-6:59" "https://www.youtube.com/watch?v=WlsJxcesWqo" --force-keyframes-at-cuts --merge-output-format mp4 -o "less-than-1-percent-domestic-applicant.mp4"

The “White Whale” of Ag-Tech

“Strawberry harvest, I call it the white whale of ag-tech in the sense that it’s something that’s chased sort of every year… And in specialty crop agriculture, 40 percent of the production costs are still labor. And in some of those crops like strawberries, it’s as much as 60 or 70 percent, in table grapes, it’s 80 percent of production costs.” — Danny Bernstein, 4:31

yt-dlp --download-sections "*4:31-5:18" "https://www.youtube.com/watch?v=WlsJxcesWqo" --force-keyframes-at-cuts --merge-output-format mp4 -o "white-whale-of-ag-tech.mp4"

The Nine-Month Gap Before Startups Touch a Farm

“We surveyed startups and we asked them this question: After you raise venture capital, how long until you had access to an actual farm? And the answer was nine months. So companies were building a deck, maybe building a prototype… and they didn’t have access to an actual farm for nine months.” — Danny Bernstein, 8:21

yt-dlp --download-sections "*8:21-9:03" "https://www.youtube.com/watch?v=WlsJxcesWqo" --force-keyframes-at-cuts --merge-output-format mp4 -o "nine-month-gap-farm-access.mp4"

Agriculture Is National Security — the “Blueberries Peru” Story

“This is national security stuff. We are importing more and more fresh produce than ever before… we’ve had a policy of subsidizing agricultural production outside the United States. So we need this technology. This is like resiliency tech.” — Danny Bernstein, 16:59

yt-dlp --download-sections "*16:18-17:29" "https://www.youtube.com/watch?v=WlsJxcesWqo" --force-keyframes-at-cuts --merge-output-format mp4 -o "blueberries-peru-national-security.mp4"

”Everybody wants to eat the pie, nobody wants to pick the apples”

“Everybody in venture’s like, ‘I’m just going to YOLO into the trillion-dollar company.’… Who’s going to do the idea of picking the apples? Everybody wants to eat the pie, nobody wants to pick the apples.” — Jason Calacanis, 18:00

yt-dlp --download-sections "*18:00-18:37" "https://www.youtube.com/watch?v=WlsJxcesWqo" --force-keyframes-at-cuts --merge-output-format mp4 -o "eat-the-pie-pick-the-apples.mp4"

The J-Trade: $100K of Figma in the Secondary

“I looked and I was like whoa, Figma is worth like 11 billion… I was like whoa, Figma’s a screaming deal. So I put in a J-trade and I bought $100,000 worth of Figma in the secondary.” — Jason Calacanis, 26:30

yt-dlp --download-sections "*26:30-27:32" "https://www.youtube.com/watch?v=WlsJxcesWqo" --force-keyframes-at-cuts --merge-output-format mp4 -o "j-trade-100k-figma.mp4"

Key Points

• Why Danny left Google for farming (2:04) - 20 years in Silicon Valley, 10 at Google; left the “period of abundance” (six messaging apps) to find where tech is needed more
• What “specialty crop agriculture” means (3:00) - Not corn/soy “big ag” but fresh produce, fruits, vegetables, tree nuts — Driscoll’s, Taylor Farms, Wonderful, Tanimura & Antle
• Root AI, the strawberry-picking cautionary tale (3:29) - TWiST-backed MIT computer-vision picker that got “caught up in SPACs” and didn’t work out
• Reservoir’s three problem sets (4:31) - Chemical reduction, chemical replacement (laser weeders, steam, UV), and automating the hardest-to-fill jobs
• Bonsai Robotics live on the farm (5:24) - Autonomous machine doing precision application of biologicals through vineyards, running a trial that morning
• The politics behind the labor gap (6:21) - Jason on the closed southern border, NAFTA-era seasonal labor, and the H-2A visa that “doesn’t scale”
• Steinbeck country (8:07) - Lon notes the Salinas setting of John Steinbeck’s novels about hard farm work
• Why ag-tech never scaled (8:21) - SF-based teams had no farmland; Reservoir cut the nine-month wait to zero, one hour from Silicon Valley
• The business model (10:52) - Reservoir Farms ($5-7M ARR operating co, 10 FT staff) plus Reservoir VC ($3M proof-of-concept fund, five checks)
• It’s a lab space, not WeWork (11:51) - 20 startups pay $3,000-$6,000/month; opened March; fund and incubator decisions kept separate
• No spray-and-pray (12:36) - Revenue subsidizes operations; hard costs include John Deere tractor leases across 100+ acres
• Aiming for profitability (13:46) - Partnering with Western Growers and John Deere to build “resilience and longevity” against ag-tech’s boom-bust
• Farm-ng became Bonsai Robotics (10:52) - One of the five proof-of-concept checks; another company being re-upped
• Agriculture as national security (16:59) - USAID-subsidized South American production created reliance on imports; domestic tech is “resiliency tech”
• The Gal Shir “AI beat me at design” tweet (21:28) - Viral post claiming Fable 5 produced a “genuinely brilliant” logo concept on its own
• The Dylan Field misfire (24:08) - Field read Jason’s late-night Paris tweet as “design is dead”; a public back-and-forth followed
• 99designs history (25:19) - Spec-work site that enraged designers (“no spec work” movement) when logos cost $3K-$25K
• AI raises the design floor (28:05) - Average design goes from ~6 to 8.5; “you can tell when somebody vibe-coded their landing page”
• The Figma bull case (34:22) - Customer base, $10B acquisition currency, driven founder — “the hat trick”; Jevons paradox for AI tools
• The moat question (33:19) - Jason asks whether Anthropic/OpenAI could out-iterate Figma; Danny: “steel sharpens steel”
• The “do no harm” earnings test (36:37) - Programs whose grads earn less than non-degree workers could lose federal loan eligibility; 800K students affected
• Danny’s tougher proposal (38:11) - Track graduates five years; if they can’t service loans, the school should be on the hook for half
• Jason prefers cost-vs-value (40:45) - Calls the policy “a hammer when a scalpel would be better”; dislikes reducing people to earnings
• The Paul Graham cheating chart (42:51) - Econ 1170: 18 students dropped, 9 skipped the in-person final once it moved offline
• Student 52 (45:29) - 64.5 on the in-person exam vs. 100 on the AI-assisted midterm — the biggest delta
• Blame the teachers (46:21) - Jason: “lazy goddamn teachers” grading AI work with AI; wants every test in person, K-12 through college
• Go analog (48:15) - Laptops closed, Moleskine and a $9 Zebra pen; writing by hand slows the brain and builds mental models
• Slow is smooth, smooth is fast (49:39) - Jason on committing to a Kurosawa film or a high-bit-rate album to counter doomscrolling
• Off Duty: Taylor Sheridan (51:24) - Lioness, Mayor of Kingstown, plus Sicario and Wind River screenwriting credits
• Off Duty: Human Vapor (53:26) - Japanese sci-fi remake on Netflix Lon calls “absolutely bananas”
• The mortality memoir (55:37) - Not Fade Away by Peter Barton & Lawrence Shames; a cable exec’s meditation on dying
• Jason’s snake grabber (58:36) - A 60-inch alloy gripper for coral snakes, rat snakes in the chicken coop, and ranch scorpions

Mentions

Companies

• Reservoir Farms (1:26) - The operating company running the 40-acre incubator; reservoir.co
• Reservoir VC (10:52) - The $3M proof-of-concept fund; reservoir.vc
• Google (2:04) - Where Danny spent 10 years before leaving big tech
• Bonsai Robotics (5:24) - Autonomous precision-application machine trialing on the farm; formerly Farm-ng
• Root AI (3:29) - MIT strawberry-picking startup that got caught up in SPACs
• Western Growers (13:46) - Largest and oldest US grower association; Reservoir partner
• John Deere (13:46) - Tractor leases and a Reservoir partner
• Driscoll’s, Taylor Farms, The Wonderful Company, Tanimura & Antle (3:00) - Examples of specialty-crop growers
• Figma (21:28) - Dylan Field’s design company; subject of Jason’s $100K J-Trade
• 99designs (25:19) - Spec-work logo marketplace that enraged designers
• Capital Factory (12:59) - Josh Baer’s Austin incubator; 1%-of-common membership model
• 500 Startups (12:02) - Dave McClure’s “$125K then take back $25K” model
• Launch Accelerator (10:25) - Jason’s own accelerator; uses Figma internally
• Adobe (34:22) - Didn’t buy Figma; whose product suite Figma could invade
• Anthropic (18:00) - Cited as the “easy” $500B-to-2x venture bet; possible Figma competitor
• NetSuite (9:04) - Sponsor read (NetSuite Next)
• Squarespace (19:25) - Sponsor read (Blueprint AI Builder)
• Y Security (30:08) - Sponsor read (on-demand security team)

Products & Technologies

• H-2A visa (6:59) - Imported farm-labor visa that “doesn’t scale”
• Lovable (32:45) - AI app builder cited alongside Figma’s AI potential
• Claude / Codex / Composer (32:24) - Coding/writing AI tools; Lon’s “you can tell when it’s Claude”
• Fable 5 (21:28) - The model that produced Gal Shir’s “brilliant” logo
• Claude Design (32:45) - Cited among impressive new design tools
• Grammarly (48:00) - Jason’s approved “supplement” tool
• Zebra G750 pen / Moleskine (48:15) - Jason’s analog note-taking kit
• Focal Utopia / Qobuz (49:39) - Headphones and hi-bit-rate streaming for “slowing down”
• Jevons paradox (34:22) - Efficiency increases consumption; applied to AI-built products

People

• Danny Bernstein (1:26) - Founder of Reservoir; ex-Google
• Jason Calacanis (0:37) - Host, broadcasting from Paris
• Lon Harris (0:48) - Co-host, in Austin
• Gal Shir (21:28) - Designer behind the “AI beat me at design” tweet
• Dylan Field (21:28) - Figma CEO who read Jason’s tweet as “design is dead”
• John Steinbeck (8:07) - Salinas author invoked on the hardship of farm work
• Dave McClure (12:02) - 500 Startups founder; early incubator model
• Josh Baer (12:59) - Late Capital Factory head; 1%-of-common model (Jason says he recently passed away)
• Nicholas Kent (36:37) - Under Secretary of Education quoted on the earnings test
• Paul Graham (42:51) - Shared the viral cheating chart
• Taylor Sheridan (51:24) - Creator of Lioness, Mayor of Kingstown; wrote Sicario, Wind River
• Zoe Saldana (52:31) - Lead of Lioness; Gamora / Neytiri
• Colin Farrell (54:30) - Star of Sugar and The Penguin
• Peter Barton & Lawrence Shames (55:37) - Authors of Not Fade Away
• Ben Sasse (57:59) - Former senator; reflective in his terminal illness
• Tony Hsieh, Om Malik, Dave Goldberg (55:37) - Late friends Danny is “meditating on”

Surprising Quotes

“It’s romanticized in the sense that people say things like, oh, you know, we have these incredible jobs for young people, people want to work outdoors… But the data would very clearly say that domestic workers do not want to do this sort of work.” — Danny Bernstein, 6:59

“We actually invested heavily through USAID to establish agricultural infrastructure in places like Peru as a deterrent to being in drugs. And what ended up happening is that they continued working in drugs, but they got significant agricultural infrastructure and created the entire South American blueberry business.” — Danny Bernstein, 16:18

“You made the easy bet on Anthropic when it was 500 billion and you got your 2x. You know what? Who’s going to do the idea of picking the apples? Everybody wants to eat the pie, nobody wants to pick the apples.” — Jason Calacanis, 18:00

“You know who’s responsible for this? It’s not the students. It’s the lazy goddamn teachers… Teachers in the age of AI are the culpable or more culpable than the students because the teachers know this is happening.” — Jason Calacanis, 46:21

“The act of writing makes you remember things and it helps you build a mental model… And hand speed, the speed of your hand is slowing down your brain. And slowing down increases thoughtfulness.” — Jason Calacanis, 48:15

Transcript

Danny Bernstein: 0:00 There are over 400,000 farm jobs that were posted last year and less than 1% of them received a domestic applicant. These are jobs that domestic workers do not want to do, that’s categorical and clear.

Lon Harris: 0:11 Nobody wants to really pick stone fruit in the height of summer in 110-degree heat, like—

Danny Bernstein: 0:16 They don’t want to do it, they need to do it because they want to provide for their families.

Lon Harris: 0:19 Exactly.

Danny Bernstein: 0:20 Strawberries, it’s as much as 60 or 70%, and table grapes, it’s 80% of production costs is labor.

Lon Harris: 0:25 Strawberry harvest, I call it the white whale of ag-tech.

Danny Bernstein: 0:28 We have basically built this space to bring agricultural technology and sort of deep tech folks into a world region to actually solve these problems.

Jason Calacanis: 0:37 All right. Bonjour, bonsoir, mon amis! J-Cal here in France, I’m in Paris. The world tour continues. Had a couple of weeks in Hawaii. What time is it for you in Austin?

Lon Harris: 0:48 It’s 8:00 AM! Our guest is in Salinas, California, so it’s about to be even earlier over there.

Jason Calacanis: 0:52 Which means it’s 6:00 AM, which means we should be respectful of our guest. Now, why don’t you introduce him and let’s get started.

Lon Harris: 0:58 Sure. Uh, we’re talking to Danny Bernstein. He’s the founder of The Reservoir. This, Jason, it’s a fascinating thing. It’s a pre-seed and seed incubator and studio for ag-tech startups, but their core product is you get physical access to a working farm. We’re going to be talking to Danny from Reservoir Farms in Salinas, California. That’s his 40-acre agricultural innovation campus. Danny, thanks so much for being with us so early in the morning today.

Danny Bernstein: 1:26 Yeah, you know, farming starts with the dawn, so here we are. It’s good to be with you guys. This is, yeah, we actually have a trial that’s about to happen this morning, so this is an early start here at the Reservoir.

Lon Harris: 1:37 Had to get up and slop the hogs and then do a podcast, you know.

Danny Bernstein: 1:41 Yeah, exactly, exactly.

Jason Calacanis: 1:43 Thanks for coming on the show, Danny. Obviously, agriculture is being not disrupted, but it’s certainly, maybe it is being disrupted. I’ll leave it for you to tell us. But how do you wind up running a gentleman’s farm there and deciding you want to do the Y Combinator or Techstars of agriculture? How’d you come up with the idea and how’s it going?

Danny Bernstein: 2:04 Yeah, it’s going well. I mean, the background, so I spent close to 20 years in Silicon Valley and I was at Google for 10 years. And when I was at Google, I saw that sort of period of abundance where the company would have multiple applications even for the same problem. Google famously had what, six different messaging apps? And so I started to think about, well, where is technology needed more? Where are there a lack of solutions for really critical problems, sort of critical imperatives? And so I left big tech and started to study complex outdoor industries and landed on specialty crop agriculture as an area that needed a lot of the same kind of principles, energy, and thinking that Silicon Valley has developed and really honed and perfected, but could that be applied to an area that needs it more?

Jason Calacanis: 2:51 Specialty crop agriculture. Specialty crop agriculture. SCA. That’s a lot of words. Yeah, but what does it mean? I mean, we can guess here, but what the heck is that?

Danny Bernstein: 3:00 Yeah, so you can kind of— Break agriculture into two buckets, but specialty crop… so you have broad acre which is like corn and soy, that’s that’s actually like big ag, subsidized ag, programmatic ag in the United States. But specialty crop agriculture is actually where our our micronutrients come from. That’s fresh produce, fruits, vegetables, tree nuts, etc. That’s high value crop agriculture. It’s the western growers, it’s the it’s the Driscolls, the Taylor Farms, the Wonderfuls, the Tanimura & Antles. So we’re basically doing technology for produce.

Jason Calacanis: 3:29 Okay, so… yeah, we had a… we had an investment in this space that was doing picking, using robot arms to to pick… it was Root AI to pick strawberries. And it was an incredible pitch. They wound up getting caught up in SPACs etc. and didn’t work out well, but pretty brilliant idea. Computer vision, computer hands, from MIT. They would know which strawberries on these vertical farms were ripe and ready to pick, and then, you know, picking a strawberry very delicate procedure but with computer vision you’re basically cataloging them all the time. And if you were going to send those strawberries, I don’t know, two days away, you might pick them a little bit more… a little less ripe, and if you were just going to, you know, use them in, you know, within a one day trip, you’d pick them a little more ripe. And this type of stuff’s been promised for a long time.

Lon Harris: 4:25 What’s taking so long, or is it just not as profitable a business, or the technology needed to catch up a bit?

Danny Bernstein: 4:31 I think the technology needed to catch up a bit. I mean, strawberry harvest, I call it the white whale of ag-tech in the sense that it’s something that’s chased sort of every year, you have two or three startups that kind of throw their hat in the ring at this. So at Reservoir Farms, we focus on three problem sets. One of them is the reduced… reducing the use of chemicals, so chemical reduction. The second is chemical replacement technologies that replace chemicals full on. So this could be electrical weeders, laser weeders, steam fumigators, using UV light to kill pests, etc. And the third category is automation of the hardest to fill jobs. And so automation of the hardest to fill jobs in that category, you would say is where strawberry picking comes in. Um, and and in specialty crop agriculture, 40 percent of the production costs are still labor. And in some of those crops like strawberries, it’s as much as 60 or 70 percent, in table grapes, it’s 80 percent of production costs.

Lon Harris: 5:18 Is labor. Yeah. And what are we seeing here on the screen here? This looks like an autonomous pesticide or a water laser?

Danny Bernstein: 5:24 This is actually a… this is a machine called Bonsai Robotics. And actually the team is over my left shoulder because they’re about to run a trial this morning over on the farm. And they have a machine that is running autonomously as you can see, in this case through vineyards, and they are doing precision application of biologicals, bio-based alternatives to pesticides.

Lon Harris: 5:44 Got it. And so this would have normally been done by a human driving a tractor of some type?

Danny Bernstein: 5:49 Yeah. It could be a human driving a tractor, it could be a human with a backpack. So that’s sort of a combination of both precision application, using less… using less chemicals… using bio-based alternatives, it also is is having someone come out of the field, which is not a desirable job. These are jobs that domestic workers do not want to do. That’s categorical and clear. The American Farm Bureau said that there were over 400,000 farm jobs that were posted last year and less than 1% of them received a domestic applicant. So we are basically…

Jason Calacanis: 6:21 All right, so you’re being very delicate there, but there is an overlay here and I’ll just say it candidly: we shut the southern border. For the last 30 or 40 years, it was open. Republicans drove this for most of that time by the way, not to make it political, because they want cheap labor for the farms. They waved everybody in and in fact, NAFTA, which people don’t even remember now, was designed specifically, I think, Danny, you and I are probably in the same age range here, I think you’re a Gen Xer as well as Lon. They remember NAFTA where they were like, yeah, just let everybody from Mexico come, work for a season then go back. It wasn’t this big polarizing thing and we’ll put all that aside. There’s no domestic workers available to do this is basically what you’re saying. This is incredibly farcical.

Danny Bernstein: 6:59 It’s romanticized in the sense that people say things like, oh, you know, we have these incredible jobs for young people, people want to work outdoors, etc. But the data would very clearly say that domestic workers do not want to do this sort of work and so we end up having these imported labor pools on a certain visa, H-2A visa, and and that just doesn’t really scale especially as we want to scale this type of production into new regions, we want to have sort of local food hubs and things like that, this type of work, it just doesn’t scale. And so what does scale in this case is the technology.

Lon Harris: 7:34 Yeah. We were talking in the pre-interview about, you know, like nobody wants to really pick stone fruit in the height of summer in 110 degree heat. Like we all love our plums and nectarines, but that’s a lot to ask. If we can automate…

Jason Calacanis: 7:45 I mean even the people who are doing it, uh, who are, you know, this is an entry-level job obviously manual labor, they don’t want to do it. They need to do it because they want to provide for their families. But this is backbreaking, hard jobs, yeah, dirty, disturbing, backbreaking labor in the blistering sun. I mean if we can retire these jobs…

Lon Harris: 8:07 I mean he’s he’s in Salinas. There’s some famous novels written by a guy from there about how difficult farm work can can be, John Steinbeck.

Danny Bernstein: 8:15 Yeah, we’re we’re here in Steinbeck country, yeah.

Lon Harris: 8:18 Alright Danny. No no go ahead. Yeah, tell us you got something else over your shoulder tell us.

Danny Bernstein: 8:21 Yeah, no right exactly. I was going to say that we we have basically built this space to bring agricultural technology and sort of deep tech folks into a rural region actually solve these problems. So a part of the, you alluded to this question of why hasn’t AgTech really scaled historically? You had teams that were based in San Francisco that really didn’t have access to farmland. So we actually, when we were building this company, we surveyed startups and we asked them this question: After you raise venture capital, how long until you had access to an actual farm? And the answer was nine months. So companies were building a deck, maybe building a prototype, putting an idea together, going out and pitching folks… It’s like you and they didn’t have access to an actual farm for nine months.

Jason Calacanis: 9:03 Crazy.

Lon Harris: 9:04 (NetSuite sponsor read) For every day your business is late to adopting AI, you fall months behind. That’s how fast things are changing. How can you possibly keep up with everything while you’re still focused on your business? NetSuite Next, of course. You already know NetSuite, the all-in-one AI-powered business management platform that connects all your data. Now with NetSuite Next, AI is built into literally every step of your process, and it’s connecting to all your data safely and securely. NetSuite Next is going to proactively make suggestions, work alongside you, and automate repetitive or routine tasks. And anytime you have a question about anything, you can just ask without interrupting a colleague. And now for the first time ever, you can try NetSuite Next for free. If your revenues are at least in the seven figures, go to netsuite.ai/twist. It’s built for every industry, ready for every boardroom. NetSuite.ai/twist.

Danny Bernstein: 9:57 It’s crazy. And so we basically wanted to figure out how we could reverse that, which is we actually now have a farm that’s here, of 40 acres. We have now multiple farms across the American West, but a startup can just give us a honk and they, as long as they’re insured, they can get onto our farm and they can start working, they can build a prototype. So that’s the idea. And in this particular one in Salinas, we’re one hour from Silicon Valley. So there’s actually really no excuse anymore.

Jason Calacanis: 10:25 Yeah, no excuses. And what it sounds like this is an open-ended invitation. You’re not doing cohorts like we do at Launch Accelerator or, you know, Techstars, Antler, Y Combinator do. This is kind of like a more bespoke thing. You’ll work with any company, any stage, and then you’ll just negotiate how you get on the cap table because everybody wants to know what’s the deal.

Danny Bernstein: 10:45 What’s the model?

Jason Calacanis: 10:46 Yeah, what’s the model? What’s the deal here? Did you start a fund? You seem like a rich guy from Google, so you’re using your own money? Do you have LPs? Is there a fund? What’s the size of the fund? What’s the standard deal? Tell us everything.

Danny Bernstein: 10:52 Yeah, sure. So basically Reservoir sits across two companies that are called interrelated. The operating company is called Reservoir Farms and that’s what runs the incubator. It’s a two-year-old company. We’ll do somewhere between five to seven million in ARR and I’ll unpack that in a second. And then we also have a fund, which is Reservoir VC. We did a proof-of-concept fund that was a three million dollar proof-of-concept fund. We wrote five checks. We are about to re-up with one of those companies. One of those companies was acquired and now became Bonsai Robotics, called Farm-ng. So there’s a proof-of-concept fund that’s operating at pre-seed and seed, and then there’s an operating company that is now 10 people full-time and is going to do over five million in ARR this year. The model with startups is they actually pay to be here. We’re operating farmland. They need to actually be good citizens and stewards.

Jason Calacanis: 11:44 So they just pay rent? This is like for your— it’s like WeWork. You just come, you pay rent.

Danny Bernstein: 11:51 It’s like a lab space. It’s like a lab space that didn’t really exist. So we have 20 startups paying between $3,000 to $6,000 a month. And we just opened in March. So that business—

Jason Calacanis: 12:00 desks, nor has anybody who rented desks.

Danny Bernstein: 12:01 Right.

Jason Calacanis: 12:02 In fact, going back 500 startups just for a little Dave McClure friend of mine, his original model was hey we’d give you 125k then he would take back 25k you would agree to spend it back with the company for office space and, you know, some other services they had. It was a little controversial on the edges, but… In other words, it’s free. I mean, 3,000 a month, 6,000 a month, I can tell you you’re not making some killing over here on rent.

Danny Bernstein: 12:20 Yeah, and that’s not our model.

Jason Calacanis: 12:21 Yeah.

Danny Bernstein: 12:22 It’s not our model because we don’t put money into these companies then ask for it back, let’s say in the form of rent. We only invest selectively, so the decision making process between the fund and the incubator is actually separate.

Jason Calacanis: 12:32 Which is great.

Danny Bernstein: 12:33 Yeah.

Jason Calacanis: 12:34 Yeah.

Lon Harris: 12:34 And I think that makes a lot of sense.

Danny Bernstein: 12:36 So we don’t kind of do that spray and pray thing of like putting a little bit of money into each. It just doesn’t pencil. And so the model basically is that we have this revenue stream, it subsidizes and supports operations. To your point, it’s not going to the moon, but it allows us to have a beginning of a relationship with these companies, and then we can start to think about adjacent opportunities with them. And that’s just beginning. I mean we’re in our second year, this is our first full year of operation.

Jason Calacanis: 12:59 Super innovative. I think my friend who just recently passed away, Josh Bear, who was running Capital Factory, had his own innovation in this model. And just to be clear, I’m not saying Dave McClure was doing anything shady with the 125k, it was kind of just innovative at the time and needed a little bit of explaining. Josh Bear did something that was super easy. It’s like, hey, join Capital Factory, we’ll take one percent of common and we’ll give you whatever, two or three years membership. And for somebody who’s just starting out, it’s like well that would have been the same cost of my cost of capital. And if you’re a breakout company, obviously it’s not your cost of capital and he would do more bespoke deals for those folks to have them in the building at Capital Factory. But that’s actually a good model for you, I think, you know, if somebody’s like $6,000 is their typical rent and you can say hey, we’ll take three thousand a month in equity, common shares, whatever, and you can pay three thousand in cash. So you know, some people might prefer that.

Danny Bernstein: 13:46 I think, yeah, it’s possible. I think maybe the difference with us is that we have hard costs. So in running a physical AI incubator, there are real costs to it. So I think that’s a bit of that transition point between sort of the fluorescent lighting, I’m in downtown Austin or San Francisco and basically I just have to kind of pump out office space. For us, we have farm lease, farm production, John Deere tractor leases, multiple farms, we’re operating over 100 acres. So we can do things like trade cash for equity, but our model has been, you know what, let’s get this thing sustainable and profitable. We’re going to be profitable next year, we’re going to be more profitable the following year. So we want to actually run this like a business. And there’s a big concern in AgTech, which you alluded to with something like Root AI, which is that a lot of AgTech sort of comes and goes. And so we wanted to think about, you know, we said early on, we’re only going to do this if we can actually see a path to sustainability. So we work with Western Growers, which is the largest and oldest grower association in the US, and then we started working with John Deere, and we kind of said to them, listen, if we’re going to do this, we gotta figure out how we can build something that has resilience and longevity. So that’s a little bit of what’s led to us so far of not doing the sort of the cash for equity thing, but that could change in the future.

Jason Calacanis: 14:54 Well, I mean, if you did a little micro fund as a test, one might suspect, I’m not having you say anything… Because I know the rules around raising funds, but Lon, I’m just talking to Lon on the side here.

Lon Harris: 15:04 Yeah, we’ll just check. Yeah, we’ll check.

Jason Calacanis: 15:05 You and I chit-chatting, and once a person does a micro fund, they might do a regular fund, and if that goes well, they might pull the string and do a macro fund, and you might just add a zero each time.

Lon Harris: 15:13 Right, exactly. Theoretically, yes.

Jason Calacanis: 15:15 Theoretically. Listen, this is a great idea, Danny. I wish you great success with this.

Danny Bernstein: 15:20 Thank you, sir.

Lon Harris: 15:21 If somebody is in AgTech, or they’re considering being in AgTech, how can they get in touch with you? How can they find out more? Because we obviously need as many people as possible pursuing important things in the world: education, construction of homes, the military, and food. Like, these are the things that are super important for America. It’s interesting that we don’t think about this as like a national security thing. It’s like, oh, all these companies are like, ‘We’re making drones and missiles,’ and I mean, having enough to eat, also very important for us.

Danny Bernstein: 15:53 Well, I mean, if you have a few minutes today, I would go to Google and look up three words, which is Blueberries Peru. And what— Blueberries. Blueberries, Peru.

Lon Harris: 16:11 Mmm, Blueberries Peru. This sounds like my next travel itinerary. This sounds great.

Danny Bernstein: 16:16 Very, very good. Highly recommend.

Lon Harris: 16:17 Okay.

Danny Bernstein: 16:18 And so what you’re gonna find basically, and probably some people watching are horrified right now, is that you’re gonna find that the US is actually played a part in subsidizing South American agricultural production, because we basically wanted to have a deterrence for South America to be doing things besides the drug trade. And so we actually invested heavily through USAID to establish agricultural infrastructure in places like Peru as a deterrent to— to being in drugs. And what ended up happening is that they continued working in drugs, but they got significant agricultural infrastructure and created the entire South American blueberry business.

Lon Harris: 16:52 Which is inc— when you think about it, the last time I bought blueberries, I can tell you, I think pound for pound it’s more expensive than—

Danny Bernstein: 16:56 Yeah, it is.

Lon Harris: 16:57 Yeah, if you go to Whole Foods or anywhere, I think it’s— the price is a little bit different, but yeah.

Danny Bernstein: 16:59 So I think what Lon is saying, which is absolutely true, is like this is national security stuff. We are importing more and more fresh produce than ever before, and it becomes really difficult for a— for the California grower, for the Michigan grower, etcetera, because we actually have a— we’ve had a policy of subsidizing agricultural production outside the United States. So we need this technology. This is like resiliency tech. This is tech that is like fundamental to like grassroots production. So that’s why we exist.

Lon Harris: 17:29 Amazing. Danny, just one more time, and you’re a humble guy, I’m trying to give you that window to get that plug in. Where can people find more? How can they contact you, Danny? Because people want to do this, I’m sure.

Danny Bernstein: 17:39 Cheers. So DMs are open. We’re Reservoir Farms across all platforms. You can visit us at reservoir.co, which is our operating company. We just launched reservoir.vc, which is a website for our fund. So you can hit us in both places, you can kind of see that separation. Yeah, please get in touch and we definitely would love to chat with you and get to know folks.

Lon Harris: 18:00 Awesome. Well done.

Jason Calacanis: 18:00 this is super innovative, I wish you great success in yours. We need more people, Danny, who are willing to do the year zero and the year one work. Everybody in venture’s like, ‘I’m just going to YOLO into, like, the trillion-dollar company and I’m going to get my, you know, 2x, 3x, you know, in three years.’ Okay, great, congratulations, you made the easy bet on Anthropic when it was 500 billion and you got your 2x. You know what? Who’s going to do the idea of picking the apples? Everybody wants to eat the pie, nobody wants to pick the apples.

Lon Harris: 18:33 Very good.

Jason Calacanis: 18:34 That’s a really relevant analogy.

Danny Bernstein: 18:37 No, you’re right. I mean, I think this is like frontier tech, this is old school and you know like getting during the boots, building a thing. And so yeah, I mean we’re seeing it live, we’re seeing the actual tech work in context, we’re bringing entrepreneurs from MIT into the Central Valley of California, and yeah, that’s what we’re doing. I mean, we’re about to run a Bonsai Robotics machine out there and reduce some chemical use. Let’s make it happen.

Jason Calacanis: 18:58 Awesome. Well, Danny, thanks so much for being here, this was fascinating and really good stuff. All right, come visit.

Danny Bernstein: 19:05 All right, great success.

Jason Calacanis: 19:06 I will come visit, absolutely.

Danny Bernstein: 19:08 Yeah, I’d like to check that out as well.

Lon Harris: 19:11 He’s not doing [beep] on.

Jason Calacanis: 19:13 Oh, no.

Lon Harris: 19:14 He’s not doing [beep]. He’s doing blueberries.

Jason Calacanis: 19:16 Blueberries. We’ll at least get some blueberries.

Jason Calacanis: 19:25 (Squarespace sponsor read) AI tools are making it easier than ever to run your own business, even as a solo founder, but you still need a beautiful, attention-grabbing website to help your new company stand out in a very crowded field. And you don’t want AI slop. Nope, you want to use Squarespace. That’s the easiest and fastest way to turn your idea into a real business, because the team at Squarespace cares deeply about design and functionality, and a plain-looking or generic or AI slop website, man, that’s going to be a red flag for your customers, for your investors, and people who want to come work for you and join your team. But Squarespace will take all the guesswork out of designing your first website with the Blueprint AI Builder, which has been finely tuned to make beautiful websites. Squarespace isn’t just going to help you make a new webpage either, they’re going to be your all-in-one platform for launching your business. They’re also going to help you set up your email, they’re going to handle invoicing, paperwork, all your needs. Go to squarespace.com/twist for a free trial, and when you’re ready to launch, go to squarespace.com/twist for 10% off your first website or domain purchase.

Jason Calacanis: 20:22 I love anybody who is doing something to support founders in that first year, I call it year zero. Because man, it’s so much hard work and the I can tell you like the number of startups that pull through onto a next round of funding when you’re this early, it’s like 50%, 60%, 40% like depending on the market conditions, if people are writing a lot of angel checks or they’re not. Hot market we had one time, seven of seven companies pull through during the accelerator at like peak ZIRP. Other times, like right after the market collapsed in COVID, we had like three of seven pull through or two of seven pull through, you know, in the accelerator. six months after it. So it’s really hard. It’s really hard to make this work.

Lon Harris: 21:03 And when you’re doing something that’s really forward thinking, that’s a little different, that’s outside the box, it can be really intimidating because you know, like if you’re making the kind of product that there’s a lot of other iterations, there’s confidence building in that. Like, obviously there’s a market for this, it’s possible. If you’re blazing your own trail, you’re alone and people are telling you you’re nuts all the time. So, a lot of pressure.

Jason Calacanis: 21:23 What else is in the news? What do we got going on here? I know people want me, there’s been a lot of stuff that’s gone down.

Lon Harris: 21:28 Yes. And you’ve been watching the news, and you’ve been watching me get in a Twitter squabble- comment on the news, sure. I think we should talk about this Gal Shir design tweet and the ensuing, ensuing kerfuffle. So, the Lemonade and Color Hunt vet tweeted, AI beat me at design. He claims a former client created a logo with Fable 5 that understood the brand story, the values, the audience, the strategy, and turned it into what he calls a genuinely brilliant concept, the kind of idea that captures everything at once. Sort of like the Holy Grail of a logo where you look at it and you’re like, that’s the perfect logo for that business. He said it’s better than what he would have come up with and Fable 5 did it all on its own. Here’s the money quote, I think. Now that AI finally took my job, I can peacefully quit and dedicate my life to studying the only thing it may never achieve, human consciousness and the pathways to God. Very, very poetic there. So Dylan Field of Figma, he responded to this. He said, I’ve never been more confident in the role of design and the impact design can have. I wish designers felt the same confidence. This is the moment to be more bold, to take more creative risk, to double down on the power of design. You then shared Dylan’s comments, I think meaning to agree with him. You said, I heard the same thing from magazine designers when Pagemaker came out, then again when 99designs and logo contests emerged. Not sure if this time is different. So my understanding is you’re saying we’ve heard this all before.

Jason Calacanis: 23:00 Every time the end of design, we’ve heard this before. Correct. Now, I wrote this in Paris after being at a bistro. There might have been a beverage, there was a cigar, there was a crepe, and it was France versus whoever they beat last night in the World Cup. So I kind of just fired this off and it might not have been perfectly crafted. I probably could have used you editing it, my bad.

Danny Bernstein: 23:24 Fair enough. It was Morocco. That’s who France beat in the World Cup. Morocco.

Jason Calacanis: 23:28 All right. Sorry you are editor’s choice. Well, Paris wasn’t burned down and people were like, hey, if France loses, independent of who they lose to, this city might be on fire. That was my so… That’s when they start throwing the furniture into the streets and lighting it on fire, Les Mis style. Exactly. I mean, they do that also if they are worked- if the government goes to a 40 hour work week from 32. They also burn the city.

Danny Bernstein: 23:51 Parisians are the best. They’re the best rioters in the whole world. We’re all following in their footsteps.

Jason Calacanis: 23:57 Yes. So here we go. We stepped in it again. I’m going to tell you. Like he’s going to do a dunk on me and he’s like oh my god JCal said this that’s the next part of the story because he misinterpreted it or I didn’t write it clean enough but I’m a fan of Dylan.

Lon Harris: 24:08 Dylan thought you were saying design is dead. I believe if a fair reading of it you were saying the opposite. But he responded never been more bullish on design in Figma than seeing Jason be negative. Looking forward to proving you wrong again. Then Jason shot back, I don’t know what the again is about. Like I’ve always believed in Figma, Figma’s great.

Jason Calacanis: 24:24 I don’t know what the again is about. Like I’ve always believed in Figma, Figma’s great.

Lon Harris: 24:27 I think he thought you were saying these people in the past were correct. That you were saying designers were like 99designs is going to kick my ass and then 99designs was this brilliant innovation that designers were not ready for. I think he thought you were taking the negative side of that. I agree you were not, but maybe it was not clear. Anyway, then you said we’re in agreement design will level up across the board as a well-designed site or app moves essentially free thanks to AI tools. But finally, the culmination of this whole tale, a J-trade happened.

Jason Calacanis: 24:59 Yes. So let me be clear here, I have been through this for a long time. So long, and this is one of the problems is I’m getting old now, I have too much history and I got to explain everything. Check this out, this is Business Insider from June 4th, 2012.

Lon Harris: 25:18 Wow, a long time ago.

Jason Calacanis: 25:19 And I had been talking about 99designs and 99designs was very interesting. It was a spec work website where you say hey I need a logo for Engadget or Autoblog or Joystiq when I was doing blogs at Weblogs Inc and I’d say the price is $250 or $500. And then you would get like 20 different choices then you would give them feedback and eventually you pick a winner. This then really tweaked designers at the time. People were very pissed off about this, that startups adopted this technology.

Lon Harris: 25:46 Right.

Jason Calacanis: 25:47 But what happened was because there were so many startups, they didn’t have the money Lon. They didn’t have the money to spend like a minimum of three to five thousand dollars is what it costs to make a logo. You know, if you were doing the next logo it might be five hundred thousand dollars back in the day, you know a million dollars for like a big company. But it was pretty common for people to spend 3, 5, 10, 25 thousand dollars on a logo design.

Lon Harris: 26:07 And I mean I think there’s a there is work that people might be listening to that and hear like oh my god, but like there’s a lot, it’s not just like draw a little logo and that’s it. It’s like a whole color scheme, it’s really like a brand philosophy design behind the whole project that you’re doing. We just went through this with This Week in AI, we sort of tried to do a whole design process for it, it can be pretty involved and yes, it will run you 5, 10K or more.

Jason Calacanis: 26:22 Well and then you might also not like the design, right? Which you know sometimes happens with a certain type of customer like me.

Lon Harris: 26:29 Right, yes.

Jason Calacanis: 26:30 So anyway, putting it all aside, I then went and I looked and I was like whoa, Figma is worth like 11 billion, not sure what their revenue is now but I did a little research on it and I was like whoa, Figma’s a screaming deal. So I put in a J-trade and I bought $100,000 worth of Figma in the secondary. …thousand dollars worth of Figma. And I tweeted like listen, people misunderstand you know have misunderstandings all the time. So you know all’s fair in love and war and on Twitter you you know text, you sometimes the tone doesn’t get down or you misrepresent something or whatever. And whenever that happens the lesson for me is to get curious. So whenever like this kerfuffle here is just over me writing an unclear tweet, Dylan not understanding what I was trying to say or maybe a little bit of both. Okay fine it happens. Either I could have written it late night, he could have read it maybe he’s on tilt because of something else.

Lon Harris: 27:32 I do think a lot of design people were like fired up about this tweet yesterday. Like I’ve saw a lot of people sharing this original tweet

Jason Calacanis: 27:41 The original tweet.

Danny Bernstein: 27:41 the ‘design is dead’ one. Right.

Jason Calacanis: 27:43 Right.

Lon Harris: 27:44 I think that put a lot of professional designers and people who really care about design sort of on edge. It’s a theme that’s sort of out there in the world that I think a lot of people are trying to sort of shut down or change that trend or the idea behind that trend. So I think that was probably part of it that it was a tense day for designers yesterday, yes.

Jason Calacanis: 28:05 Yes. And they are by the way when this 99designs thing went down that was the same exact thing. All the designers got really upset and then they started a movement: no spec work. So there was a movement trying to convince designers not to do spec work. Now if you’re a young designer, young designers used to email startups like Uber or Robinhood or whatever and they’d be like ‘hey I redid your logo’. This was a tried and true way to get customers. Especially if you were an offshore designer because this is like the humans in Manila, India, learn the design tools. They stole a lot of design, they did derivative stuff, you know it wasn’t the sharpest stuff. But they would just email you just like clippers do now, they would just email you a couple of samples and say ‘hey if you like my work would you give me $300 for it’. This is called spec work. And I said you know spec work is great if you’re coming up in your career. But designers hate spec work. They hate people working for free. Now that’s an individual’s choice if they want to do it or not. I’m fine if people want to do it, I’m fine if people don’t want to do it. It’s personal freedom here. Anyway, design, the big picture here is, is design going away? The answer is no. What’s happening though is the average website, the average app I would say 10 years ago I would rate as like coming out of a startup in years one and two was probably like a six or a seven out of design if calm.com and Robinhood were a 10. For example if you looked at the first version of Uber it was like a six or a seven. It was clunky. Now when they started making money and they raised money they eventually were like a 9.5 or a 10 out of 10 in terms of how beautiful the app was. But they did it over time. Now I think using the AI tools and you know Squarespace did this to websites right all these beautiful templates they had, you know the average went up. So the average design is going to go from a five six or seven to a seven eight nine using these tools. I’ll say

Danny Bernstein: 30:00 You know, 7.5 to 8.5 is going to be the average and I would say the average used to be like a six. So that’s like a significant jump.

Jason Calacanis: 30:07 Yeah.

Jason Calacanis: 30:08 (Y Security sponsor read) The team is pumped because you’re about to close a massive deal. But then the client’s lawyers get involved. What happens if you get hacked? How are we going to protect your data? There’s no need to panic. This is why you brought in Y Security. Y Security is staffed with over 40 experienced engineers who have actually worked security for world-class companies like Apple, Uber, Microsoft, Robinhood, Brex, and so many others. But the best part is you don’t even need to hire Y Security. Your company can rent Y Security’s elite team by the hour. That means no massive salaries to pay, no costly consultants, just real experts embedded in your company helping you out with your SOC 2, ISO 4200, or any security or compliance challenge you’re facing. You can even set a monthly cap so you know exactly how much you’re spending. And your first six hours are completely free. Head to ysecurity.io/twist and book your free six-hour strategy call. That’s ysecurity.io/TWIST.

Danny Bernstein: 31:06 However, when you do have money, you do wind up hiring a designer because you say, ‘Hey, I want to stand out from the average.’

Lon Harris: 31:25 Right. Yeah.

Danny Bernstein: 31:26 So the average keeps moving up so you’re going to have less ugly products. Just like you don’t have slow websites anymore. This used to be the major issue, like, ‘Oh, your website’s not optimized and it’s slow.’ Now all websites are fast. Like, when’s the last time you used a product or service like, ‘This is slow’?

Lon Harris: 31:31 Yeah.

Jason Calacanis: 31:32 Yeah, it’s not an issue.

Lon Harris: 31:33 Yeah. I think, I think you’re right. And I think that you can already see it so much, like, there’s so many, you know, tweets that are clearly written by Claude. And they’re not, they’re not terrible, but when you read a really well-written post or an article, it stands out because you’re like there’s so much sort of slop going on. This person who clearly sat down to write something thoughtful themselves, it actually is even more valuable. I think it’s the same. Like, you can tell when somebody vibe-coded their landing page. It looks, it doesn’t look bad, but it looks like how Claude or how Codex or how one of these programs makes a landing page.

Jason Calacanis: 32:03 I think that’s fair enough, yeah.

Danny Bernstein: 32:05 And but, and then, but because the things are getting better, I’ll be honest, the 8-8.5 that I said, I think like next year it’s going to be like a 9-9.5. So it’s going to start, and you can prompt it, ‘Hey, do better, do better, do better or I’m unplugging you,’ you know.

Lon Harris: 32:21 Right. Sure, yeah.

Jason Calacanis: 32:24 But most people don’t. I think that’s what’s, I mean, to me, this is a whole separate conversation. But I think what’s holding back AI and in terms of the public perception is that most people don’t go to that second level of like iterating, putting themselves in the loop, making things incrementally better with help from Claude or Codex or Composer or whatever they’re using. They just take the first output and go, ‘There it is, done. Copy-paste.’

Lon Harris: 32:44 Yeah.

Danny Bernstein: 32:45 I bought some Figma. Why did I buy Figma? I think right now, they’re going to, I think they’re going to figure out AI. And if you look at Lovable and you look at Claude Design and like what these are capable of doing now, it’s pretty impressive. already has all the designers and they’re going to just they you know Dylan understands this customer base better than anybody. They are trading at seven times sales, they are trading at a pretty high PE like 70X, but I think their business is going to get accelerated by AI and I think they’ll be one of the top three or four players in the space no matter what. So I placed a bet. So I placed a bet.

Jason Calacanis: 33:19 I did have one question. I mean what my my only reservation here. I think Figma is great. We use Figma internally at Launch for all sorts of stuff. It’s it’s we all like Figma. It’s no knock to Figma. But I guess my question would be are you not concerned there is a moat issue there where it’s like I mean all of the frontier labs have their own design products they’re all thinking about this as a category like is there a chance that Anthropic and OpenAI just come up with the next iteration of what Figma would be and steal Figma’s thunder?

Danny Bernstein: 33:50 Um I think they will have more competition and by having more competition what does it steel sharpens steel?

Jason Calacanis: 33:58 Yeah there you go.

Lon Harris: 33:59 And I think I guess it’s I think I’ve heard both versions.

Danny Bernstein: 34:03 Iron sharpens iron I’m not sure what the term is here but I think you know this new-found competition that they have is going to make them sharper. The blade’s going to get sharper. They also have a currency…

Lon Harris: 34:14 It is iron sharpens iron it’s from Proverbs 27:17 I’m being told.

Danny Bernstein: 34:19 Yeah I was just reading Proverbs just the other week.

Jason Calacanis: 34:21 Oh there you go.

Danny Bernstein: 34:22 So iron sharpens iron I think that they also have a currency as a ten billion dollar company to go buy other companies and teams. So I do not look at Dylan and Figma as like oh my god they’re in a terrible position. They’re in a great position. Designers love it. And so as they keep adding AI features as they go acquire customers and they’ll be able to move into other adjacencies. So since Adobe didn’t buy them I I would suspect Figma could go into all of the Adobe suite of products and you know because of how efficient it is to go into adjacencies and that’s the trend in AI. So if you have a customer base if you have a currency to buy companies which they have both of those and you have a driven founder which they also have is three for three that’s the hat trick they’re going to figure it out. That’s my belief they’re going to figure it out. And Jevon’s paradox as we have these AI tools we’re seeing this internally we’re seeing it in the startups who apply people are going and building more products and services they’re going right down the line of products and services that weren’t worth building because it was too niche and now they’re building them so uh I think it’s going to have great success.

Lon Harris: 35:40 Jevons Paradox by the way uh when technology makes a resource more efficient or effective you actually see use of it go up not down. It’s based on steam engines right steam engines cause coal to go used to go up not not down was the original Jevon observation. And as it gets cheaper.

Danny Bernstein: 35:57 Not financial advice but I put a J-trade in.

Jason Calacanis: 35:59 There you go fair enough.

Danny Bernstein: 36:00 That’s what it is. At—and—and let’s have Dylan back on This Week in Startups, of course.

Jason Calacanis: 36:01 Should we talk about…

Danny Bernstein: 36:02 I know set a reminder set a reminder every three months for me to check in on this.

Jason Calacanis: 36:03 Yes. No, we just did a flashback the other day. We featured you and Dylan from the early days of COVID, talking about Figma.

Danny Bernstein: 36:10 That was incredible because we were in the office during the early days of COVID debating like what would happen. I was like, ‘Yeah, this is all going to be over in April, May. I’m going to be able to do my next accelerator class.’ I had no idea what was happening. He was like, ‘I think there’s going to be a second wave.’ I was like, ‘Ah, who knows? Stay home for two weeks, we stop the surge, we’re good.’

Jason Calacanis: 36:26 No one knew in mid-March in fairness to you guys. No one, no one knew. Alright, we should talk about this—you want to talk about this colleges article you sent me?

Danny Bernstein: 36:36 Yeah, let’s do lightning round on that.

Lon Harris: 36:37 All right. The US Department of Education is rolling out a new accountability test for colleges and universities. Here’s the test: if graduating students don’t earn more than comparable workers who didn’t go to college, the program that they attended could get cut off from federal student loans. Same would go for masters programs. This is the one that I think is more controversial. The same would go for masters programs from which grads earn less than someone with just a bachelor’s degree. Key quote here from the Under Secretary of Education, Nicholas Kent: ‘If a program cannot show that it leaves its graduates financially better off than if they had never enrolled, it should not be underwritten by federal taxpayers.’ Critics argue that college, and especially grad school, they’re not just about earning money. We need nurses and teachers, but what about artists, musicians, filmmakers, all of those kinds of people, not guaranteed to be locked into a great salary when they graduate? A few more stats from the Department of Education: 800,000 students they estimate currently attend a program that would fail this test. 18% of undergraduate certificate programs would also fail this test. So I guess to me the big question here is: what do you think about this policy, and how much of college and student loans for college should be about ‘you’re going to take this degree and get a good higher-paying job than you would have otherwise’ versus what we used to think of college as being about, which is it’s this time for growing and learning who you are and finding a community and all that other non-payment related stuff?

Danny Bernstein: 38:11 Yeah, so taking the two questions. Number one, this doesn’t go far enough. This should be much more cutthroat, but they’re obviously starting somewhere. And the test really should be: are you able to pay the American taxpayers back or not? And if you’re not able to pay them back, then the school should be on the hook. That’s my belief. And the way to look at this is they should just look at school by school. How many people graduated from this program? Let’s say it’s design. You got a design degree and you took a federal loan. So you went to Parsons Design School or RISD, whatever it is. And you track that person for the five years after they graduate and they have their loan. Do you have a job? How much money are you making? If they are not making enough to service their loans… Then, that program should not be eligible for, uh, these federal loans. Why? You can go spend whatever you want out of your pocket, but we’re talking about the taxpayer’s money here. This is you and I and everybody else in America deciding to give a loan to somebody to go get a philosophy degree, then a Master’s in philosophy.

Jason Calacanis: 39:23 Right.

Danny Bernstein: 39:24 And what that will do is it’s going to make these colleges lower the price of those degrees, which is what you really want to do. These degrees should be priced to perfection.

Jason Calacanis: 39:29 Right.

Danny Bernstein: 39:30 If you’re going to go be a dentist and you want to spend $300,000 or $400,000, you want to be an MBA and you’re going to spend $200,000, $300,000…

Jason Calacanis: 39:41 Right.

Danny Bernstein: 39:42 …okay, we know in those career paths that there is a job waiting in all likelihood. In these other ones where there’s not a job, why do they cost as much? Like, the Master’s degree should not for art cost the same as the one for being a doctor or for being a lawyer. It’s so obvious. Should those things exist in the world? Of course. But these colleges need to control costs. The federal government has a way to stop this, and I think we just need to really change, we have to put it on the colleges. If they want to have their students get federal loans, they should be on the hook for half of that loan. If that person doesn’t get a job or that person defaults on it, the college should have to pay half of it. That’s what I would do, and I think that’s the direction we’re going in. So this is long overdue, and the reason why these colleges can charge $250,000 for an undergraduate degree or a graduate one and put people in this massive debt is because there’s no backstop. They just keep taking more government dollars and they keep raising the price, and they’re lying to students that this degree’s going to be net positive.

Jason Calacanis: 40:41 Right.

Danny Bernstein: 40:42 It’s not.

Jason Calacanis: 40:45 I like your take on it more than the policy that’s being proposed, because your take is tied between how much are you paying for this degree versus how much is its real-world value, which I think is more like a comparison. This is kind of saying, do you earn more than somebody with a bachelor’s degree? Well, not everybody who goes to grad school is even looking to necessarily out-earn anyone who has a bachelor’s degree. There’s all kinds of reasons you might want to go to a Master’s program, and that doesn’t mean you wouldn’t be able to eventually pay your Master’s degree back. I think this is like using a hammer when I think a scalpel would be better. I mean, obviously, we all want to bring the cost of education down, and I do think this is a real problem, that we’re graduating so many people with degrees in fields where there’s, you know, 30 jobs, and it’s like, well, what is everybody else with that degree supposed to do?

Lon Harris: 41:30 Yeah, everyone’s gonna go to… I mean, is everyone gonna go be a director and a screenplay writer? Like, there’s not a lot of jobs, folks.

Jason Calacanis: 41:37 Film school is always like the stark example, where yeah, like every once in a while you get a George Lucas, but that’s not every person who graduates from film school. A lot of people just like movies. So, you know, I don’t… the one thing that I don’t like is when we like reduce people and college and studying and all of these fields to just, well, how much do you make? Like, you’re more… you are more than how much you make, I think, as a person. And I don’t like that part of the…

Danny Bernstein: 41:57 Listen, I have a better proposal, and by the way, there’s…

Jason Calacanis: 42:00 Probably five other better proposals, but at least, at least we’re starting to address the issue. A bad proposal is a precursor to a better proposal.

Lon Harris: 42:08 Right. Well, and that’s always how the Trump administration tactic is just throw something out there, now you’re having a conversation. Your original idea doesn’t have to be the one that gets through. You’re just… it’s an opening volley and I think that’s how you always have to understand all of these.

Jason Calacanis: 42:25 But they gotta ship. You know, you can say whatever you want about the Trump administration, I’ve got plenty of criticism of them.

Lon Harris: 42:31 And I do.

Jason Calacanis: 42:32 And you do?

Lon Harris: 42:33 And I do.

Jason Calacanis: 42:34 And any reasonable person might. But they do ship.

Lon Harris: 42:38 They ship product.

Jason Calacanis: 42:39 You’re absolutely right. They ship product. Sometimes too much product. Sometimes really weird product.

Danny Bernstein: 42:46 Donald Trump moves units, I don’t think you can—I don’t think you can argue that. Sure, let’s make a ballroom. Sure, let’s make a new Air Force One. Sure, let’s keep going.

Jason Calacanis: 42:51 But let’s focus on what matters. Education is a critical piece. And by the way, all of this is deck chairs on the Titanic. The real issue is, how do we use AI to help people get skills? There was another really interesting thing that Paul Graham tweeted, you can go find it here. He—there was a study that he showed. In this study, it was—a midterm occurred and everybody, it seemed like, was using ChatGPT and had like aced the midterm. The teacher ran it through like this software, the software gives a lot of false positives, etc. to check for cheating. He’s like, ‘Okay, I think y’all cheated. I think like 90% of the students cheated.’ And then he said, ‘I’m going to do an in-person, we’re going to do an in-person…’ and they show this chart. So they do an in-person final. So between the two, you have the scores of the midterm and then you have the scores, and here it is. Zoom in on this.

Lon Harris: 43:32 Mhm.

Jason Calacanis: 43:33 So, here we go. These are, in gray on the left—zoom in a little bit more so we can actually see it, just like give us the first the top part of it, please, so we can see that. Okay, so Econ 1170 midterm and final scores. After Serrano made the final in-person—Serrano’s the teacher—18 students dropped the class and nine students didn’t take the exam. The scores of the remaining 59 students are displayed here. So look at student number one. They scored 95 on their midterm, which they did on their own, and on the in-person, they got 95.5. In other words, there’s no gap. And then let’s go down and look at everybody else, and you start to see this huge gap, right?

Lon Harris: 44:18 Mhm.

Jason Calacanis: 44:22 And then, there’s my favorite student, S22. Midterm, he didn’t cheat.

Danny Bernstein: 44:37 But he’s not that bright either.

Lon Harris: 44:39 They just didn’t do well.

Danny Bernstein: 44:41 But at least he’s honest.

Jason Calacanis: 44:44 That’s the guy you hire. That’s the gal you hire. They’re not the smartest student in the class, but—

Lon Harris: 44:49 They’re gonna try their best.

Jason Calacanis: 44:51 And then scroll down and you start to see, like, there are people, you know, here who obviously are cheating. 7:58, 59, I think. They didn’t even take the test. I’ll scroll back up. The gray dot. Look at how many people scored a 100. Scroll all the way up. So, I think I may have juxtaposed it. The midterm score is the orange, Final exam score is the one in gray. That’s the one in person. So, when you see these scores drop off, who’s got the widest here? There’s a bunch of people with a 100. If we scroll down, let’s find the fattest person. Who’s got the biggest delta?

Lon Harris: 45:27 Oh my lord, who is that?

Jason Calacanis: 45:29 Oh, come on now. No, look at student 52. Student 52. Shout out 52.

Danny Bernstein: 45:36 He got a 100. This mofo, this fella’s hard work. 64.5 on his own, 100 with the help from AI.

Jason Calacanis: 45:43 By the way, everybody’s cheating.

Lon Harris: 45:45 Well this is, it’s tied exactly into what I was saying. I think, I think…

Jason Calacanis: 45:50 This is like work from home.

Lon Harris: 45:54 People are just not using these AI tools well. Like they’re not, we didn’t educate people into the back and forth that you have to do. Everybody who I know who uses these tools effectively, it’s a conversation. It’s back and forth. You’re saying no, that’s not what I wanted or I don’t like this or rewrite this or are you sure about this fact or rewriting things in your own voice. Like it’s a tool. It would be like going to Google and just copying and pasting a Google result into your test. That’s not how you do it. It can be a great help, but you have to like be engaged and it’s not a shut off your brain thing and I think that’s how people are-

Jason Calacanis: 46:21 And here’s the thing. I don’t know if this was at Oxford or where this was, but it’s at a significant college. And by the way, they all gave this guy ish for bringing it up. Right. You know who’s responsible for this? It’s not the students. It’s the lazy goddamn teachers. These teachers don’t want to even review the work. I can tell you firsthand. Teachers in the age of AI are the culpable or more culpable than the students because the teachers know this is happening.

Lon Harris: 46:50 The students are writing their projects in AI. They’re doing AI slop as proven clearly here.

Jason Calacanis: 47:02 Yes.

Lon Harris: 47:03 The teachers on the other side are using AI to review it and then yolo, everybody gets an A plus. The teachers are lazy, the students then are following the teachers down the primrose path, is that the right term?

Jason Calacanis: 47:16 Yes, sure.

Lon Harris: 47:17 The path of evil.

Jason Calacanis: 47:18 They’re following them. Their example. The teachers need to be fired. If you’re doing anything work from home, anything on your own, I don’t care, it doesn’t matter. I want every single test in person. No, nothing at home. I am infuriated about this even in primary education. Students are doing this right now in fourth grade through 12th grade in college. I want everybody…

Lon Harris: 47:45 Did you use blue books when you were in school?

Jason Calacanis: 47:47 Sure, oh yeah, of course.

Lon Harris: 47:50 Scantrons and blue books, that’s what it was when I was in school.

Jason Calacanis: 47:53 But go back to old school. Get your pen and paper out. I want to see you compose a sentence.

Danny Bernstein: 47:59 Now when you get into…

Jason Calacanis: 48:00 The real world, fine, you don’t have to have penmanship, you can YOLO it, use ChatGPT, use Grammarly. I love Grammarly.

Danny Bernstein: 48:05 It would be nice if people did.

Lon Harris: 48:06 It would be nice if people used these tools as supplements. That’s what I think they’re supposed to be. Like I like Grammarly. Grammarly’s helpful, but it’s not…

Jason Calacanis: 48:15 What do I do at the management team meeting? I tell everybody: laptops closed, have a pen, and have a Moleskine and be adult. And I want you to write down during the meeting, this is when we have our management team meeting, I want you to write everything down in your Moleskine. You want to take a picture later? You want to look at the Zoom transcript? It’s all good, no problem with that. But the act of writing down makes you present in the meeting. And what you write down is what you as a human believe is important, and that matters as well. So we have to keep the analog going. We have to keep our brains engaged. I have, within arms length somewhere here, I’m not going to get it, but I have my journal. I have my Zebra G750 $9 f’in pen that writes incredibly smooth. I turn off my stuff when I’m on a podcast, when I’m doing a meeting, I write notes myself. Why? The act of writing makes you remember things and it helps you build a mental model.

Lon Harris: 49:10 Yeah, it’s just how your brain works. It’s the actual act of writing it out in your own hand helps you remember and connect ideas together.

Jason Calacanis: 49:21 Yeah.

Lon Harris: 49:22 It’s just like when you have something to memorize, you should read it right before you go to bed and then go to sleep and you’ll wake up knowing it better because that’s how your brain works. And we’re just disconnecting ourselves from all of this stuff and letting farming out our thought to AI. And that’s not, I think, what we should do. It’s a supplement, it’s not a replacement. That’s what people don’t…

Jason Calacanis: 49:39 And hand speed, the speed of your hand is slowing down your brain. And slowing down increases thoughtfulness. So what you’re doing actually is you have a regulator. The regulator on your brain is the pen and paper. Now, if you take your brain, and our brains move fast, and you turn on a microphone and you start just recording stuff and spewing stuff, and then have AI spewing stuff, you’re just going a thousand miles an hour, nothing remains in your brain. It’s the same as doomscrolling. When you go sit and commit in a movie theater to turning off your goddamn phone and giving yourself over to a two-hour Kurosawa film or a Paul Thomas Anderson film, pick your favorite director, it lets you slow down to the actual input. You can watch the movie, hear the dialogue at the speed of your ears and eyes can process it. Slow down to speed up and understand and enjoy this. I had to do this for appreciation of music. When I’m feeling scattered, I put on my Focal, like, incredible headset. Shout out to my friends at Headphones.com and Focal. I put on my Focal Utopias, and I listen to my Qobuz French streaming service… at the highest bit rate and I will slow myself down to appreciate an album at the highest bit rate at the and and just enjoy that experience. And I do it on cigar time. I light up my cigar, I got my book, I’m taking notes, I’m listening to the music, I’m slowing myself down. It’s like meditation and you will be more thoughtful.

Lon Harris: 51:18 Yeah. It’s that it’s that military saying slow is smooth, smooth is fast? It’s like that same concept.

Jason Calacanis: 51:24 Correct! That is the saying. I by the way, I I was looking for something to stream. I love Tyler Sheridan. I’ll I’ll kick it off here. There’s a thing called Lioness. Sure. It’s a CIA show he’s doing, Nicole Kidman, Morgan Freeman, I know. Incredible. And it’s it’s it follows the playbook of Tyler Sheridan. This stuff just moves. Tyler Sheridan is retard max.

Danny Bernstein: 51:46 A little bit. A little bit.

Jason Calacanis: 51:47 No, I’m serious because he just goes with his gut. It’s like mile a minute energy, it’s uh, you know, really uh intense stuff and he just goes there and like uh oh my god did he did do Sicario?

Lon Harris: 52:06 Yeah he wrote he he wrote the script for Sicario. It’s the Denis Villeneuve directed but Tyler Sheridan wrote the wrote the script. He also did this really great movie called Wind River that you should check out.

Jason Calacanis: 52:13 Oh okay well there’s your off duty. So anyway I am loving this show I’m on the second season Lioness is about recruiting women specifically, thus Lionesses, for this CIA program I will leave it at that but the the two or three women who are the leads in this…

Lon Harris: 52:31 If you’re Zoe Saldana, yeah, she’s like the main lead, isn’t she?

Jason Calacanis: 52:36 Zoe Saldana, I don’t even know who this person is but she is a tour de force.

Lon Harris: 52:40 She’s the the main Na’vi lady from the Avatar movies and she’s Gamora, the Guardians of the Galaxy she’s the green lady Gamora.

Jason Calacanis: 52:46 Oh she’s Gamora, okay. Well who knew who knew. Anyway, she is incredible, the show is incredible, just like Landman you have these great actors and you go on an adventure it goes a million miles an hour. Is it like high art? I don’t know.

Danny Bernstein: 53:02 No he’s got he’s really good at setting up these scenarios these like intense you know rivalries or whatever and then just playing them out he he’s very good at what he does. I really like that one.

Lon Harris: 53:11 There’s a lot at stake. There’s a lot at stake in these shows.

Danny Bernstein: 53:12 He does that one Mayor of Kingstown too for Paramount Plus with Jeremy Renner that I really like.

Lon Harris: 53:19 Oh I gotta watch that.

Danny Bernstein: 53:21 Yeah he’s like a fixer in a town that’s run by private prisons in Mayor of Kingstown.

Lon Harris: 53:26 All right there is a Japanese series right now on Netflix called Human Vapor that is absolutely bananas and I’m I’m really enjoying it it is a it’s like a remake of an old like 70s series I think and they just did it in this really goop it’s got it’s a combination of a lot of stuff it’s like a sci-fi fantasy premise where there is this guy who’s literally made of vapor who’s committing all of these crimes and so the cops are trying to find out who is the human vapor what’s happening so there’s a conspiracy level but it’s also this really high concept sci-fi premise. But it’s also kind of funny and goofy. It’s really in that—in that way that only the Japanese would do. Like, Americans would find this show kind of all over the place, and “What’s the tone supposed to be?” and “How am I supposed to feel about this?” The Japanese just go for it, and I really admire the—the creativity, the originality. I got really sucked in; I’m, like, five episodes into this now. But—

Jason Calacanis: 54:19 This looks great. I, you know, I need a new sci-fi thriller, so this looks like a great recommendation.

Lon Harris: 54:23 It’s a lot of fun. Uh, yeah, Alice. And if you’ve been to Tokyo, like as I have, it’s fun to watch a Tokyo—this is the first show I’ve watched set in Tokyo since I visited there.

Jason Calacanis: 54:27 Hmm. Okay. And what else you got? Anything else?

Lon Harris: 54:30 Oh, uh, I was just going to say, uh, Sugar is back for a season two on Apple TV+ with Colin Farrell. I don’t know if you watched this, Jason.

Jason Calacanis: 54:43 I have not. I don’t even know what you’re talking about, which is why I love you.

Lon Harris: 54:46 It’s an—it’s an LA Noir detective series. Colin Farrell plays a private eye named John Sugar, who’s investigating missing persons cases in LA. There is a—there is another level to what’s going on in Sugar that I will not dream of spoiling. So check it out. Towards the end of season one, a bit of a—a bit of a surprise twist that upends the premise, but if you like LA detective Noir mystery shows, I think you would probably enjoy it. Sugar on Apple TV+.

Jason Calacanis: 55:06 I like the actor. I like the actor. Uh, it’s interesting to see him bounce back and forth between this and The Penguin, because Sugar is, like, sort of at heart, like, a really good guy who’s trying to help people. It’s a very heroic, sort of kind-hearted role. And then The Penguin, he’s—he’s the exact opposite; he’s a real—a real slimeball.

Danny Bernstein: 55:37 So this book is, um, really, really interesting. It’s about an executive, Peter Barton, who—I didn’t know him personally, but I knew of him because I came of age during the cable industry boom. And the cable industry was known—like, it was kind of like the AI industry of its time or the mobile/dot-com boom. And this guy just made hundreds of millions of dollars, which, like—he made $200 million, like, when that was a lot, a lot, a lot of money. And then he finds out he’s got stomach cancer. And he writes this book in his last year and a half of life with this guy, Lawrence Shames. And it really is, like, a meditation on dying and what matters in life. And given that my friend Om Malik passed away, my friend Josh Baer passed away, and then a little earlier, you know, Tony Hsieh passed away from Zappos, my other friend, and Dave Goldberg from Launch.com, the music service, passed away, and SurveyMonkey—I’ve been really meditating on, ‘Hey, life, and how to have a good life.’ And this book—it nails it. And I think if this—and it’s not a long book, but what’s really interesting in terms of a device, the guy Lawrence, who is the co-author of it—he’s, like, a professional writer—he, in between the chapters written…

Jason Calacanis: 57:00 by Peter Barton, he talks about his time with Peter Barton. He only knew Peter Barton during this last 18 months of his life with cancer diagnosis. But then he got to know everybody who knew him during his life. And he gets to compare the person, you know, everybody’s opinion of this person, Peter Barton, during his career,

Lon Harris: 57:20 Yeah.

Jason Calacanis: 57:21 versus when he post-cancer diagnosis and he’s dealing with, he’s got young kids. What’s what’s going to happen to my daughter or my son, you know, they’re 10 years old and they’re going to say goodbye to their dad this year. And, you know, trying to get closure for all of these things. Anyway, um, if you’re into meditating on a life well lived, this is a good book for you to do that. If you’re young and invincible, you don’t need to read these things. But I would recommend it. If you’re young and invincible and on top of the world, at some point you’re going to deal with your own mortality and you’re going to want to start thinking about this issue of what is a good life? What is it worth focusing your energy on?

Lon Harris: 57:59 It makes me think of Ben Sasse, the the former Nebraska Republican senator. He he is dying of cancer right now and he’s been he’s been doing lectures and he’s sort of I’ve seen him do interviews and you do, like, when you get towards the end of your life with one of these terminal illnesses, it just changes you and it it makes people more reflective and it it softens their their sharp edges in some ways. So I’ve been seeing a lot of interviews with him lately that are the same thing where he’s he’s getting reflective in his final months.

Jason Calacanis: 58:29 Well, and you also become super…

Lon Harris: 58:31 …candid.

Jason Calacanis: 58:32 Candid, because you’re going to be gone, so you’re kind of like, well, yeah, let me just tell you the truth.

Danny Bernstein: 58:34 People got to hear about this now while I have time.

Jason Calacanis: 58:36 Exactly. Um, and here’s my second thing. This is a tool. Uh, people may have seen this like coral snake that was on the ranch and uh it was injured and I had to put it down. But I then had a…

Danny Bernstein: 58:51 Oh my god.

Jason Calacanis: 58:52 rat snake on the ranch. So between the coral snake and the rat snake I bought this 60-inch professional alloy gripper snake grabber. The wife and the kids tried to get the rat snake out of the chicken coop. Now, they’re not there to eat the chickens, chickens are too big for this, but the the eggs. They really like…

Lon Harris: 59:11 The eggs.

Jason Calacanis: 59:15 Really, you need to have this if you’re in on a ranch. People don’t tell you when you move, Lon. I don’t know if you’ve had the scorpions in the house yet.

Lon Harris: 59:22 I have. Yeah.

Jason Calacanis: 59:24 Okay.

Lon Harris: 59:24 These and you gotta be careful with your shoes. Just always give them a little tap out because people put their shoes on or their slippers and they get a nice little stinger from a scorpion.

Danny Bernstein: 59:32 I I had the large - I had a huge scorpion in my bathtub one morning when I woke up to go to the restroom. I didn’t even know they got that big. That’s crazy. It’s crazy.

Jason Calacanis: 59:44 No, they - it’s not going to kill you, but it’s - it is going to pack a punch and it’s not going to be pleasant.

Danny Bernstein: 59:48 It’s gonna suck. Yeah, it’s - you’re not going to enjoy yourself.

Jason Calacanis: 59:51 Anyway, just have a little bit of a joke there. Alright, listen. Another amazing episode of This Week in Startups, your favorite podcast if you’re a founder or you’re an investor. We got a lot of great shows for you over the summer and I will be here for most of them. Alright, I had to

Lon Harris: 1:00:00 As I was in transit a little bit so apologies I wasn’t here, but I’ll see you next time. Bye-bye.

Jason Calacanis: 1:00:04 Bye, buddy.