Announcing 2075.ai - embracing the AI renaissance

Today’s a very special day as we introduce 2075.ai to the world. 🎉

TL;DR:

2075 is an innovation lab that brings together builders and domain experts from around the world to build new and impactful products and ventures.

To build new ventures, we follow a simple approach:

1. Identify industry problems that we’re passionate about solving

2. Assemble, guide, and fund small teams of exceptional builders and domain experts to solve said problems

3. Develop and test products in the market and with customers, and seek validation–or lack of it

4. Spin out successful products as standalone ventures, retaining some equity and providing ongoing support

We’re launching our first venture, Expert.Med, in early 2026. Expert.Med is a healthtech platform that allows patients to get a fast, accurate, and inexpensive second opinion on their medical imaging data from qualified doctors, assisted by AI. 

If you’re looking to build a venture together, please follow this link to get in touch.

In a time of unprecedented potential and necessity for innovation, in addition to building new ventures, we also offer an innovation-as-a-service product to enterprises that are looking to boost their in-house disruptive innovation efforts. Get in touch here if you’d like to learn more.

That covers most of the what and the how. If you’re curious as to why we started 2075, please read on.

Venture building

If you ever built a software product before ChatGPT, Cursor, and Claude Code came along, you probably remember that it used to take 3-10x more resources to get a new product off the ground. You would normally raise a bit of money, hire 5-10 people, and get to work (as I did when I built Aylien). 

AI has drastically changed this, and in most cases, if you have a deep enough understanding of the problem you’re trying to solve, a good engineering mindset, and good taste, you can create products at a level of quality where software isn’t your limiting factor (as opposed to other factors such as distribution, a good understanding of customer needs, etc), for a fraction of the cost.

Venture success ≈ product × distribution × operations × funding

Product success ≈ working software × good UX × strong user–problem fit

Understanding user needs ≈ your first-hand experience with the problem × market research

Interestingly, AI hasn’t directly impacted distribution, operations, or funding, anywhere near as much as it has affected product development. 

Moreover, historically, most founders I’ve come across seem to fall into two camps:

1. Have worked in an industry and encountered a hairy problem or set of problems that they set to solve with their startup.

2. Have not worked in an industry, and were therefore building software to solve someone else’s problems–YC seems full of companies like this these days.

In other words, unless you had first-hand experience with a problem, you would essentially be brute forcing the market until you hit a nerve. This is another area where I feel AI can have a huge impact: what if we could be more systematic about how we identify, quantify, and prioritize industry problems, using data, AI, and input from human experts?

Following the trajectory of these two AI-accelerated dimensions–easier software development, and a systematic approach to problem identification–to the asymptote, you get something that isn’t too dissimilar to how a quantitative trader operates: 

A quant will:

1. Algorithmically identify opportunities (e.g., a security is under- or over-valued)

2. Craft a strategy and deploy capital to take advantage of the opportunity

3. If they were right, the market rewards them

So, similarly, a software business builder should be able to:

1. Look at various data sources and analyze them to identify and size opportunities

2. Deploy capital and have a swarm of AI agents build a software product to take advantage of the opportunity

3. If they were right, the market rewards them

That’s setting aside distribution and operations, which are equally hard problems, and most likely the primary limiting factors in this new post-AI world. However, given enough time, AI will undoubtedly diffuse into these as well. 

So the core thesis I have for 2075 is that if we:

1. Systematically identify pockets of opportunity where demand outweighs supply,

2. Bring together domain experts who understand a space really well, and builders that have embraced AI-first software development,

3. Build and launch MVPs to validate and test our assumptions (desk-side research only takes one so far), and

4. Apply our fundraising, GTM, and general venture building expertise,

we can build new products and ventures that create value, which is shared between the parties involved in creating each venture.

One could say that in this regard, we’re somewhere between a single product startup, and a traditional startup accelerator like YC (many investments, alignment to thesis, hands off).

Innovation-as-a-service

Most successful innovation hubs throughout history have benefitted from stable, reliable funding that allowed the innovators to explore new and ambitious ideas with greater freedom. Be it a cash-rich ad business in the case of a company like the early Google, a rich parent company like AT&T for Bell Labs, or wealthy and generous patrons in the case of the Renaissance.

Meanwhile, a classic Innovator’s Dilemma dynamic is now visible across the industry where the best companies are hyperfocused on the needs of their current clients, potentially leaving them exposed to the rapid, AI-driven disruption. The need for exploring new and disruptive solutions to existing or emerging problems, without risking the core propositions and revenues is a crucial part of the enterprise strategy playbook in 2025. 

2075.ai’s innovation-as-a-service offering gives enterprises a way to explore new ideas without distracting their core teams, while supplying us with a cash flow to allow us to continue to innovate. We work closely with product, strategy, and domain experts to understand the organization, the market and the opportunities in front of it. Then, using the same ideation, prototyping and validation approach we apply to our own ventures, we generate a steady flow of tested MVPs that can be piloted with customers and brought into the roadmap. It’s a practical, low-risk way to stay ahead of the AI-driven shifts in the market while keeping the core business focused on day-to-day execution.

If you’re interested in exploring an innovation-as-a-service collaboration, please get in touch.

Looking ahead

For any creative builder, this is one of the most exciting periods in the history of humanity.  The limits of what’s possible have vastly expanded, with plenty of whitespace to explore and create value around.

2075’s aim is to bring together forward-thinking builders and give them the space, tools and support to explore new ideas and turn them into real ventures. We believe the next wave of progress will come from small, curious teams working fast, learning fast, and caring deeply about the problems in front of them.

If you’re a builder, a domain expert, or someone sitting on a problem worth solving, we’d love to hear from you. The future isn’t going to build itself, and there has never been a better time to get started.