Thesis

We all consume huge amounts of information online, but often don’t know the source. The idea was to create a certificate/digital signature for media, so that wherever it shows up, consumers (platform and end users) would be able to ask for the media source and better assess its trustworthiness.

The problem is huge and will continue to grow since Gen AI is creating huge amounts of information. This has led to significant concerns among content companies/media outlets as well as tech companies, but people don’t know what to do. There is a sense that there needs to be a third party to fix it – it has to work across platforms, and many of the current players have conflicting interests. A third party can increase the trust between the players and lift all boats.

We started with video, since there are tens of thousands of videos made per minute. We developed a cryptographic signature that verifies that the video is from a specific source, including per frame, in case the video format is changed. The technology also allowed us to detect clips – e.g., this video is a pure subset of a longer-form video that has a known source. This capability was designed to address the market concerns about snippets of interviews that were taken out of context. We patented it. 

Validation

During the first 6 months we spoke to a lot of senior people – 30-40 people across media/publishing houses, GenAI platforms, distribution channels, regulatory experts, etc.

• Top media outlets: We had conversations with leaders (CPO/CDO/CISO) from top global brands. They are all concerned, but don’t really know what to do. In retrospect, we didn’t speak to enough influencers/smaller players. 
  

• Platforms – Distributors/Social Networks/GenAI – That’s where the issue largely lies, since much of the content is posted across platforms/accounts and not directly by the creator channel. We found that they are interested and concerned and working to address the trust/safety/quality challenge. They are using internal and external resources to try and detect content quality/truthfulness/harmfulness. It’s assumed to be a cat & mouse game, of constantly beating the algorithm, and the technology we were offering could help all players.

Why the idea didn’t progress

1. Our solution can only work at scale, but not clear what value you bring to customers on Day One. 

2. We assumed news is the most painful vertical to address, but it’s a market without money

3. Fund raising:

• Given our strong track record, we wanted to make a big syndicate raise ($8-10M) that would allow us to launch with big publishers.

• We learned that some investors didn’t want to touch media at all (previously burned). Many others liked the big vision and long-term strategy, but weren’t convinced we were solving a significant enough problem on Day

4. Lack of urgency: One of our advisors connected us with industry leaders who work on related problems through a consortium. Their experience is that our target audience of content creators are not moving as quickly as they thought, and expect it will happen once regulators get involved. But once there is regulation, it will become a commodity.

5. Potential market failure: a question we got is why this hasn’t been developed yet, given that the basic technology for media signature and trust challenge has been around. The sense was that there might be dynamics/market failures that we don’t see, like maybe big interests are at play.

Insights

1. It’s a topic that a lot of people are interested in, so it’s relatively easy to get to conversations with people in the industry. 

2. Start by talking to 100-150 customers to identify those who would be willing to participate and only then go to investors. We thought we’d be able to invest money first, given how big the idea is.

---