Building a Large Geospatial Model to Achieve Spatial Intelligence

Occasionally we turn over the Partner PoV column to our entrepreneurs and portfolio companies. Today we turn it over to Niantic. This is an incredible article describing how Niantic builds giant alternate worlds in cyberspace. It’s fascinating!

Building a Large Geospatial Model to Achieve Spatial Intelligence

At Niantic, we are pioneering the concept of a Large Geospatial Model that will use large-scale machine learning to understand a scene and connect it to millions of other scenes globally.

When you look at a familiar type of structure – whether it’s a church, a statue, or a town square – it’s fairly easy to imagine what it might look like from other angles, even if you haven’t seen it from all sides. As humans, we have “spatial understanding” that means we can fill in these details based on countless similar scenes we’ve encountered before. But for machines, this task is extraordinarily difficult. Even the most advanced AI models today struggle to visualize and infer missing parts of a scene, or to imagine a place from a new angle. This is about to change: Spatial intelligence is the next frontier of AI models.

As part of Niantic’s Visual Positioning System (VPS), we have trained more than 50 million neural networks, with more than 150 trillion parameters, enabling operation in over a million locations. In our vision for a Large Geospatial Model (LGM), each of these local networks would contribute to a global large model, implementing a shared understanding of geographic locations, and comprehending places yet to be fully scanned.

The LGM will enable computers not only to perceive and understand physical spaces, but also to interact with them in new ways, forming a critical component of AR glasses and fields beyond, including robotics, content creation and autonomous systems. As we move from phones to wearable technology linked to the real world, spatial intelligence will become the world’s future operating system.

What is a Large Geospatial Model?

Large Language Models (LLMs) are having an undeniable impact on our everyday lives and across multiple industries. Trained on internet-scale collections of text, LLMs can understand and generate written language in a way that challenges our understanding of “intelligence”.

Click here for more…..

Why We Invested In Fathom AI

At Frontier, we’ve established themes that drive our areas of investment interest.  Additionally, we use tested business frameworks to identify companies which have the highest likelihood of success and where we can add the most value, based on our unique strengths as a firm. We love to back founders and companies who are uniquely positioned to transform existing industries and disrupt the status quo.

That’s why we’re excited about Fathom AI, our latest addition to the Frontier portfolio.  Fathom AI provides an AI-powered text analytics solution for businesses.  Text analytics, also known as Natural Language Processing (NLP), plays a crucial role in extracting insights generated in large volumes of unstructured data.  Fathom’s software platform transforms open-ended text responses into actionable insights, making it more efficient to interpret qualitative data at scale.  The company has generated traction with customer profiles, including consumer-facing brands, tech companies, consultancies, and market research agencies. In brief, wherever there are users providing feedback, Fathom can help find the signal in the noise! Our conviction in this opportunity was driven by a few key factors:

Team: CEO Michiah Prull and COO Tovah Paglaro are multi-time founders, building a solution informed by years of experience solving this pain point for customers.  From a technical perspective, both co-founders have a deep understanding of the quantitative foundations and qualitative inferences underpinning its technology.  And yet, they’re also able to set company strategy, execute against plan, and nurture long-term customer relationships - this a compelling skillset!  At Frontier, we believe a company takes on the culture of its founders.  We found this team credible, resolute, and focused, with the poise and maturity needed to build trust with clients of any size.

Total Addressable Market (TAM): It’s estimated that 80-90% of the data an organization generates is unstructured, the majority of which is text-based.  Within this repository of language and data, there’s a treasure trove of knowledge to be gained – opportunities for process optimization, improved customer relations and retention, user feedback to drive product roadmaps, etc.  With this potential for impact, it’s easy to understand why the Text Analytics market is forecasted to grow past $10 billion in 2024 and $50 billion by 2029, as businesses large and small look to gain new insight from their data.  Today, Fathom is targeted to community-generated data, such as customer feedback, public opinion, and employee sentiment.  While these use cases represent a massive category on their own, there’s been increasing adoption across industries and new applications emerging along the way.  With an extensible platform like Fathom’s, the opportunity for delivering value to customers grows even larger.

Technology:  Fathom’s platform leverages a blend of computer intelligence and human curation.  Its technology was developed over 5 years while partnering via closed API with OpenAI, Microsoft, and Anthropic, and combining this learning with the company’s own models, built on tens of millions of proprietary data points.  This AI-driven insight is augmented further by Fathom’s own personnel, experts who supervise the application of models to ensure the highest level of accuracy, detail, and actionability.  As a result, businesses can pinpoint critical issues affecting customer satisfaction and churn, allowing for segmented data analysis to better understand their customer base.  It’s a novel approach that’s optimized for accuracy, cost, and scale.

In summary, Fathom AI provides an opportunity for Frontier to invest in the future of unstructured data analytics.  This innovation is powered by computer intelligence, natural language processing, and advancements in artificial intelligence – as well as a rock star leadership team!  We’re thrilled to invest in their next chapter of growth.

Do The Thing.....

The Draper venture family (from l to r) Tim, Billy, Bill, Adam and jesse

The Draper family is legendary.  They are the true OGs of the Silicon Valley, starting with General Draper, and then innovated by his son Bill Draper, then his son and my partner Tim Draper.  And three of Tim’s four children are VCs!!!  The family knows the venture capital game as well as anyone.  I just read a post by Adam Draper, who founded and leads Boost VC, a firm focused on science fiction technology companies (it’s really cool, check it out). 

Anyway, his post really resonated with me.  I’d sum it up with, ‘perfection is the enemy of good enough’.  Or, as Adam says, “do the thing.” 

Click here to read Adam’s post…..

737's and Wake Up Calls

Today we turn over the keys to PPoV to Karan Talati, CEO of our portfolio company First Resonance….

Karan Talati, CEO, First Resonance

🚨 The recent Boeing 737 Max 9 incident, where a door detached due to what's been called a 'one-off quality issue', really brings something home for those of us in the business of making things that matter. These 'one-off' issues aren't just blips on the radar. They're wake-up calls, reminding us of the ever-present need to be vigilant in industries where there’s no room for error.

🔧 At First Resonance, we're not just observers; we're active participants in shaping a more resilient future in manufacturing. We believe in the power of:

- Putting the right tools in the hands of those who make things happen – the engineers and the shop floor heroes.
- Giving them the keys to the kingdom with APIs and dynamic data platforms to democratize access to information and improving quality.
- Empowering them to continuously model, analyze, and sniff out potential issues before they become real problems.

🌍 Our partners, the forward-thinking companies we work with, are already on this path. They're embracing flexible, data-driven ways of building things. Automated systems, real-time data analysis, machine learning – these aren't just buzzwords; they're tools that are being wielded right now to make everything from electric planes to the next big breakthrough in renewable energy. It’s exciting to see the transformation as these companies shift from the 'way we’ve always done it' to 'what’s the best way to do this now.'

⚙️ We're at a crossroads where embracing modern technology isn't just about staying competitive; it’s about ensuring safety and pushing the boundaries of what’s possible. Whether it’s air travel, energy, or the next giant leap in space exploration, the mantra is clear: innovate, integrate, and empower.

🚀 So, let's not just chalk up incidents like the Boeing one to bad luck. Let’s use them as a rallying cry to build better, smarter, and safer. It’s about rolling up our sleeves, getting our hands on the data, and really diving into what makes things tick. The future is now, and it’s ours to shape with every line of code, every data point, and every bold new idea. I am optimistic about a better future, with more hardware, delivered with more reliability than ever. Let’s get to work for a better future.

Visit First Resonance by clicking here…..

May The 4th Be With You!

Reposting this Partner PoV, authored by Scott Lenet in 2016, in celebration of today’s Star Wars holiday ;)

---

Darth Venture

What VCs can learn from the management style of the Sith

At my prior firm, my DFJ Frontier co-founder David Cremin coined the term “Darth Venture” in the early 2000s in one of his all-too-frequent clever moments, so I suppose we bear some responsibility for propagating the rhetoric that VCs are like the Sith, the champions of the Dark Side.

While I personally identify with (young) Obi-Wan Kenobi, sometimes you just need to lean into the dominant discourse in your industry and accept how others view you. With that mindset, I thought it would be interesting to explore whether there is anything constructive that venture capitalists can glean from the Dark Side.

Click here for more…..