The time is right for Photo Intelligence

Hi everyone, I’m Ted. If you’re reading this newsletter, we’ve probably already met and decided that it’s worth keeping up.

But for those of you who don’t know me, I’ve spent the past 20 years wearing all sorts of hats in tech — from individual contributor roles at early-stage Google and Square to leadership roles up to CTO. I'm currently working in venture capital at TheGP, evaluating investment opportunities and helping to accelerate the success of the great startups in our portfolio by embedding with them in a variety of full-time roles.

This newsletter is where I’ll be sharing my journey to solve real human pain points at the intersection of AI, photos and videos. My goal is to keep you in the loop on what I’m learning and building, spark conversations, and (when the time is right) explore ways to collaborate — whether that’s building together, investing in the vision, or simply sharing feedback.

A bit of background

At the end of January, I wrapped up a project with Thatch, an amazing company in the healthcare technology space. If you’re in the business of providing health benefits to your employees, check them out; the ICHRA model that they support is truly innovative and they have an amazing product and team.

After that, the amazing partners TheGP gave me the freedom to explore my own interests, and after researching, prototyping, and discarding a variety of other ideas, I ended up focusing on Photo Intelligence — the idea that AI can supercharge photos, enabling businesses and people to do things that were previously considered impossible, incredibly time consuming, or too expensive.

I’ve always been interested in photography. In high school, I did most of the photo editing for my high school newspaper. I bought my first digital camera in 2000, followed by a variety of other compact digital cameras, DSLRs, and mirrorless cameras. I’ve worked with photos and videos in Photoshop, Premier, Lightroom, Picasa, Google Photos, and so on. At absolutely no point have I been truly satisfied with the products available.

In early April, I reached out to Ben Lang, who runs the fantastic Nextplay network, to let him know I was starting something new around using AI to help people organize and do more with the photos and videos they take. A week later, while on Spring Break with my family, my phone wouldn’t stop buzzing — Ben had featured me in his newsletter and over 50 founders, builders, and investors reached out to talk about their pain points, use cases, and ideas.

Since then, I’ve been splitting my time between product discovery and building. I’ve had the privilege of speaking with so many passionate and insightful people about their use cases, pain points, and ideas for products in the space. There’s clearly a lot of pain, especially in workflows around curation of photos and creation of digital and physical goods from photos. There’s clearly a lot of opportunity as well.

What I’ve learned so far

The time is right for “Photo Intelligence”

Over the past few years, advances in the academic research and commercial infrastructure for analyzing and working with images have unlocked capabilities that were previously exclusive to companies like Google and Meta, enabling others to integrate more sophisticated image features into their products.

Among other things, AI and ML can now be used to:

• Index and search vast photo libraries by text or image

• Cluster similar shots together and score them for technical or aesthetic quality

• Recognize and group photos of the same person or pet

• Transform images via text prompts (e.g. “Studio Ghibli–style edits”)

• Merge multiple imperfect group photos into one idealized shot

• Enhance low-res or blurry faces using other images of the same subject

• Chat directly with your photo library (e.g. via MCP for your photo library)

And every capability I just listed applies equally to videos.

The pain points are real

Over the past few months, every person I talked to has shared their own use cases and challenges, and a few common themes emerged. Many of the most painful stories I heard involved taking a large collection of photos — whether of a person, or from a trip or event — then manually reviewing each photo to select a much smaller subset, individually editing, rotating, and cropping those photos, and only then being able to do what they intended from the start.

Some things I heard:

• “The pain is real when it comes to organizing everything. There’s so much untapped potential in helping people better organize, rediscover, and interact with their photos and videos."

• “I never do anything with those photos or videos but I know they have value.”

• “As someone with 147,000 photos in a Lightroom catalog, I can understand the need for some AI help in this area. I’m pretty diligent about adding keywords to my photos, but it’s a task that requires dedication, so there’s got to be a way to use AI to help. And don’t even get me started on video organization"

• “Smart albums would be useful as a starting point. Things like recipes, meals to start. Then it should be able to do more. For example, it could suggest a restaurant based on my preferences.”

From all these conversations, I learned that many people are spending tens or even hundreds of hours manually reviewing, organizing, and editing photos — often followed by many more hours actually creating something meaningful from them.

AI could automate much of this toil:

• What if, after your wedding, you could send each guest a thank you card automatically personalized with a photo of them at the wedding? AI could help pick the best photo of each guest for you.

• What if you could chat with AI to design a custom yearbook for your organization, based on the people, events, and themes you want to highlight? AI could lay out the spreads according to your directions, crop and adjust photos for you, and present you with a draft.

• What if, at the end of a vacation, you automatically received an album of every photo your guide or tour group took of you during the trip? AI could automatically curate and organize all of the tour photos by person or family group.

There are other examples where the time cost isn’t as high, but are still difficult enough that people don’t do them as often as they would like.

For example, I heard:

• "It would be great to keep the 25 people who matter in my life up to date through regular photos."

• "There are pretty good solutions for private photo organization and public sharing. The open space is private photo sharing and automatic curation."

What if the digital photo frames you’ve gifted automatically stayed up-to-date with the best pictures you take together?

What if, at the end of each day, an app suggested the best photos from your day to share with the most important people in your life, and suggested the same to them in return?

I believe that AI could help us maintain our most important human connections by making the simple things we want to do but don’t find time for easier than before.

I could build this

These types of capabilities are still difficult to build and integrate into products. Unless you have an engineering team with a background in photography and computer vision, expertise in applied AI, and access to sufficient GPU resources, many of these capabilities are still out of reach, and no one has yet made them accessible to “ordinary developers” or consumers.

Over the past two months, I’ve prototyped many of the techniques above and built the confidence that I could create both consumer apps that solve these pain points, as well as a developer platform that makes it dramatically easier for others to do the same.

I’ll talk more about what I’ve built so far in the next newsletter.

Where I’m headed next

While I love the consumer use cases people have shared with me, I’m currently leaning more toward a B2B/platform model. I believe that the platform model would create the leverage to have a much broader and deeper impact overall, and I would be so excited to see developers and consumers use the platform to supercharge existing products with Photo Intelligence, vibe code new tools for their own particular needs, and find exciting applications of the technology that I would have never dreamed of.

My goal over the next few months is to continue refining my understanding of the pain points and building conviction in a particular product direction. I’ll do this through a combination of conducting additional interviews, building a demo I can show to potential customers, and releasing an SDK that others can build on. I’ll aim to provide a regular update in this newsletter, so let me know if there’s anything else you want to hear about!

How you can help

• Know people with deep experience in this space, either in customer-facing or technical roles? I’d love an intro.

• Know a business with photo-workflow headaches that they want to solve, or new opportunities they’d like to go after? I’d love an intro.

• Want to build on top of a photo-intelligence platform? Let’s chat.

Don’t have my email? Just comment below or send me a DM.

Special thanks

Thank you to everyone I’ve talked to over the past few weeks.You’ve all been so helpful and insightful, and I want to make some special callouts:

• Ben Lang: for featuring me in his newsletter, resulting in this outpouring of interest and really solidifying my conviction and direction

• Chian Gong: for coining the term “Photo Intelligence”, which really cut to the core of the platform value proposition.

• Spencer Adams-Rand: for introducing the idea that a platform that helps customers hide photos of particular people (e.g. for compliance or privacy reasons) could be as important as a platform that helps customers surface photos of particular people.

• Jaron Gilinsky: for crystallizing the product gap between AI that’s good enough for a proof of concept, and AI good enough to trust with curating your photos.

• Jeeyoung Kim: for helping to flesh out product and business models that involve storing photo and video assets vs. just analyzing them and throwing away the originals, and how both are the right model, for different use cases.

• Sergei Sorokin: for his support and collaboration as we both explore our founding interests.

• Andrew O’Neal: for his insights into B2B vs. B2C business models. Like him, I’m intrinsically more interested in solving the consumer pain points, but believe the B2B direction creates the leverage to have a much broader and deeper impact overall.

— Ted