The checks being cut to ‘owners’ of training data are creating a huge barrier to entry for challengers. If Google, OpenAI, and other large tech companies can establish a high enough cost, they implicitly prevent future competition. Not very Open.

Model efficacy is roughly [technical IP/approach] * [training data] * [training frequency/feedback loop]. Right now I’m comfortable betting on innovation from small teams in the ‘approach,’ but if experimentation is gated by nine figures worth of licensing deals, we are doing a disservice to innovation.

These business deals are a substitute for unclear copyright and usage laws. Companies like the New York Times are willing to litigate this issue (at least as a negotiation strategy). It’s likely that our regulations need to update ‘fair use.’ I need to think more about where I land on this – companies which exploit/overweight a data source that wasn’t made available to them for commercial purposes do owe the rights owner. Rights owners should be able to automatically set some sort of protections for at least a period of time (similar to Creative Commons or robots.txt). I don’t believe ‘if it can be scraped, it’s yours to use’ and I also don’t believe that once you create something you lose all rights to how it can be commercialized.

What I do believe is that we need to move quickly to create a ‘safe harbor‘ for AI startups to experiment without fear of legal repercussions so long as they meet certain conditions. As I wrote in April 2023,

“What would an AI Safe Harbor look like? Start with something like, “For the next 12 months any developer of AI models would be protected from legal liability so long as they abide by certain evolving standards.” For example, model owners must:

•  Transparency: for a given publicly available URL or submitted piece of media, to query whether the top level domain is included in the training set of the model. Simply visibility is the first step — all the ‘do not train on my data’ (aka robots.txt for AI) is going to take more thinking and tradeoffs from a regulatory perspective.
• Prompt Logs for Research: Providing some amount of statistically significant prompt/input logs (no information on the originator of the prompt, just the prompt itself) on a regular basis for researchers to understand, analyze, etc. So long as you’re not knowingly, willfully and exclusively targeting and exploiting particular copyrighted sources, you will have infringement safe harbor.
• Responsibility: Documented Trust and Safety protocols to allow for escalation around violations of your Terms of Service. And some sort of transparency statistics on these issues in aggregate.
• Observability: Auditable, but not public, frameworks for measuring ‘quality’ of results.

In order to prevent a burden that means only the largest, well-funded companies are able to comply, AI Safe Harbor would also exempt all startups and researchers who have not released public base models yet and/or have fewer than, for example, 100,000 queries/prompts per day. Those folks are just plain ‘safe’ so long as they are acting in good faith.”

Simultaneously our government could make massive amounts of data available to US startups. Incorporate here, pay taxes, create jobs? Here’s access to troves of medical, financial, legislative data.

In the last year we’ve seen billions of dollars invested in AI companies. Now is the time to act if we don’t want the New Bosses to look like the Old Bosses (or in most cases, be the exact same Bosses).

Updates

• My friend Ben Werdmuller riffs on what he calls ASCAP for AI, the need for a standard licensing framework and payment structure.
• Someone also reminded me that if you care about content being valued correctly over time that you should also care about competition among AI models. That an oligopoly might not lead to higher value for content given smaller number of bidders.

Winter Break week for my kid more time with her, and when she’s with her friends, more time with New Yorker magazines. Here are a few essays, articles, blog posts, etc that I’ve enjoyed recently.

What Happens When TikTok Is Your Marketing Department [David Segal/New York Times] – Was it organic? Was it spon con? Was it both? Many times we’ll never know, but the random products that end up popping because of a TikTok trend are always pretty fascinating anthropological stories. Here the focus is on Pink Stuff, a British cleaning paste, which was #CleanTok mainstreamed to a quadrupling of revenue ($125m annually) and distribution to 55 countries.

A typical #CleanTok video features a so-called “cleanfluencer” — some have more than one million followers — working over a sink, or a pan, or a floor, with a particular cleaner and a particular brush. There are usually before and after images, which make these little vignettes a cross between a commercial and an episode of “Law & Order.” They start with a mess and end with a verdict.

The Five Lessons That Have Guided My Career [Avni Patel Thompson/Milo] – Derived from a talk she gave at a High School Career Day, CEO/entrepreneur Patel Thompson thinks that guidebooks are better than roadmaps when it comes to career advice.

Founder Vesting [Jared Hecht/USV] – Jared joined USV earlier this year and it’ll be interesting to see how his writing changes as he adds ‘institutional VC’ to his founder and angel investor knowledge. Here he writes about a topic (vesting cycles) that often is incorrectly positioned as ‘founders vs investors’ but actually has a lot more to do with the commitment founders want to make to one another and to their company. As Jared notes,

To hedge against this predictable outcome, more founders should adopt longer vesting cycles for themselves and the earliest (big equity) employees. Stretching things out to a six-year vest helps to prevent co-founder abandonment. Equally important, it also protects you if your co-founders aren’t the right fit early on – you don’t want someone leaving two years into building your company with the lion’s share of the cap table. That sucks for everyone.

The Introverts Have Taken Over the US Economy [Allison Schrager/Bloomberg] – tldr: it takes a lot more to get people to leave the house these days for dining, shopping, entertainment and other Out of Home activities. Maybe it’s just another version of barbelling? Where we like Uber Eats delivered meals but also the Eras Tour? The mediocre middle of inconvenience for little reward is getting squeezed? There’s been a lot written about decreases in IRL socializing, which has really harmful consequences for people IMO, so this could also just be a correlation/byproduct of that trend.

Enjoy!

I read a lot of stuff and here’s a few worth passing along to you!

China’s Age of Malaise [Evan Osnos/New Yorker] – A loooong read but essential stuff if you are interested in China from an sort of view (cultural, economic, geopolitical, startup).

When I return to China these days, the feeling of ineluctable ascent has waned. The streets of Beijing still show progress; armadas of electric cars glide by like props in a sci-fi film, and the smoke that used to impose a perpetual twilight is gone. But, in the alleys, most of the improvised cafés and galleries that used to enliven the city have been cleared away, in the name of order; overhead, the race to build new skyscrapers, which attracted designers from around the world, has stalled. This summer, I had a drink with an intellectual I’ve known for years. He recalled a time when he took inspiration from the dissidents of the Eastern Bloc: “Fifteen years ago, we were talking about Havel.” These days, he told me with a wince, “people don’t want to say anything.” By the time we stood to leave, he had drained four Martinis.

Incentives and the Cobra Effect [Andrew ‘Boz’ Bozworth/Facebook] – So I don’t know if the story Boz references here is fully accurate or has taken on some metaphorical expansion, but it’s worth sharing. A quick post about the power of incentives – and how they can sometimes backfire. The title is explained in the opening paragraph:

When Delhi was under colonial rule it suffered from an excess of venomous cobras. To curb the population the government paid a bounty for dead cobras. This triggered entrepreneurs to start breeding cobras to collect the bounty. When the government figured out what was happening, they discontinued the bounty which meant all the cobras being bred were worthless and were thus set free, increasing the cobra population significantly.

It’s Never Been More Important to Understand Your Capital Provider’s Business Model [Charles Hudson/Precursor] – Charles is just a wonderful human and his posts about venture capital are essential for anyone who considers themselves part of the startup community. Similar to Boz’s essay earlier, this one too is about incentives. And how mismatched (or unspoken) ones in venture capital can cause stress.

 If you are a founder and you are experiencing new or renewed tension in your conversations with your VC investors, it’s worth re-examining whether you all have a shared view of the likely outcome of your company and whether you’re both as excited about what that outcome means. In many cases, I’ve seen situations where there are founder-acceptable outcomes that are below-the-line outcomes for VCs, and that conversation goes unsaid or unexamined. This creates a lot of unspoken and unexamined tension in the founder and VC relationship.

Don’t Talk To Your Competitors’ Investors [Chris Neumann/Panache Ventures] – I generally agree with what Chris writes here although if he has experienced that most VCs share information received with a company directly with a competitive company in their portfolio that makes me sad. We try upfront to disclose any conflicts and if we, during a pitch process, find out that the presenting company is competitive with an existing investment would never forward materials.

One disconnect between founders and investors is sometimes the definition of ‘competitive’ and founders will push back to actually suggest they’re not competitive with an existing investment. I get it, they want to keep as many doors open as possible for funding. The reality is though, that especially at seed we tend to give our existing founders a very wide berth, even if it’s just adjacent. Why do we do this even at the cost of passing on an interesting company? Well we all know that early stage companies do a lot of exploration in their problem space before settling on the exact product. Even more importantly we want to be able to bring just one of the startups to their next investors as representing ‘our investment in [market x].’ I find that our conviction will strengthen the interest of Series A VC, versus us having a set of similar looking companies.

Five Traps for Real Estate Tech Entrepreneurs [Brad Hargreaves/Thesis Driven ($)] – Brad’s a multiple time proptech founder/investor and his newsletter is well worth the subscription price if you’re at all interested in real estate investing – both the technology and property holding company side. Google presents a snippet of the five mistakes so I don’t feel like I’m violating his paywall by sharing here 🙂

Hope you enjoy these as much as I did!

Great interview with surviving cast members of Happy Days, which was a number one TV show itself, as well as producing FIVE spinoffs. Worth reading the whole interview but one particular part stood out for me involving the producer Garry Marshall and star Henry Winkler.

WINKLER He [Marshall] was generous but also was structured. He took no bad behavior. One time, when he was announcing the guest cast, I said, “Garry, we have to hurry up because I’m flying to Arkansas.” He nodded, put down the microphone, grabbed me by my shirt, put me against the wall and said, “Don’t ever do that again, because they have every right to be recognized like you.” He kept us in line.

Talk about setting a tone! How often as leaders do we let little things slide or fail to apply values consistently? Holding aside the physicality of this encounter, which is very 70s and very Hollywood, it’s a reminder to me about building teams and supporting your people. And those lessons get handed down through generations.

It was a relationship, and an appreciation, that Winkler carried with him going forward