To build something truly great, you have to start with a great idea. How do you find a great idea? By being insatiably curious.

The questions our members tackle are early signals of where great ideas are emerging. If any of these resonate, connect with the SPC member listed below.

Even better, if you're wrestling with any of these questions, apply to join SPC directly.

We don't have requests for startups, we have requests for curiosity. Here's our Winter 2025 list.

What if scientific knowledge was no longer codified in published papers?

Scientific journals were an enormous advancement. In 1665. Modern global science should move at lightning speed, but it’s rife with issues. The primary way knowledge is disseminated, via peer-reviewed journals, underpins all these problems.

• Can we develop a system that better recognizes and rewards genuine innovation, wherever those contributions come from?
• What does a “paper” look like if it’s published in real time by any researcher?
• Can a new vision of the “paper” serve as the nexus for a genuinely rich, threaded, multi-party, multimodal, attributed discussion?
• Can AI synthesize all this knowledge and data, surfacing insights, hypotheses, and inconsistencies for researchers to test?

Reach out to Ruchi, Mark, John

What new business models are lurking around the corner?

Each new technological paradigm brings with it a requisite shift in the economic structure. With the internet, we saw the rise of massive social marketplaces with gravitational network effects. With mobile, we saw the gig economy shape an entirely new way of transacting online, and with cloud, we saw many shifts in how businesses can totally rethink their cost structure.

• Is selling “work” rather than “software” the best way to deliver and extract value?
• Are there novel ways to monetize a large consumer audience without relying on ads?
• What is the future of the “ad” in a world where agents work on behalf of people?
• What are the second-order effects of a “token” economy where customers pay for intelligence on-tap

Reach out to Gopal, Adam

What are the new form factors on which humans will spend their attention?

AI slop is everywhere. Humans are getting overwhelmed by the constant stream of multi-modal content coming their way. It’s becoming harder for them to figure out the best use of their time and attention.

• How will AI mirrors utilize the user embeddings and preferences to surface the most entertaining content and experiences for humans at all times?
• How will NPCs and PCs engage and create with each other?
• As boredom with the familiar goes up, what will be the new emergent form factors of speculative plays and games of chance?
• How will AI integrate into physical experiences?

Reach out to Prateek, Apurv

How do we bring AI to the next billion people?

OpenAI is closing in on a billion active users. But the next billion AI users will look very different. They will interact with technology in contexts and geographies that we haven't yet designed for. We’ll need to rethink everything, from how we prompt models to what interfaces we build.

• What does AI look like when voice, visual, and gestural inputs are primary instead of text?
• What latent signals (location, images, audio context) can become implicit prompts that eliminate the need for explicit text input?
• How do we design for users underrepresented in training data?

Reach out to Ankit, Dheemanth

What kinds of hardware and software advances will make AI compute more scalable and sustainable?

Massive CapEx spend on data centers, GPUs, and power generation reveal the limitations of the current AI hardware stack, well before anything like AGI. Combined with growing public backlash, you would expect growing pressure for innovation.

• What new opportunities do the economics of data centers (rare earth inputs, construction, permitting, etc.) create?
• What AI infrastructure will people be fine with, and what will face regulatory barriers?
• What’s the next great compute paradigm?
• Are there novel environments (the tundra? space? the moon?) better-suited for new kinds of compute infrastructure?

Reach out to Jonathan, Christian

What if we make machine learning continuous and distributed?

Today’s version of machine learning is based on outdated research. It’s been advanced by leveraging massive computation and data scaling, rather than embedding human-designed knowledge. The architecture we built was optimized for scaling, but it's time to rethink those assumptions.

• How can we improve learning speeds?
• How do we find new data to train machines, and how do we specialize intelligence for specific domains?
• How do we understand how machine learning learns?
• Are GPUs the future of computation, and how do we remove current bottlenecks (memory, power, and network bandwidth)?

Reach out to Marco, Kushal

How will we adapt to the growing attack surface created by ubiquitous data capture and the malicious use of AI?

Pandora’s box is open. The manpower constraint on sophisticated digital hacks is about to disappear as bad actors get access to highly capable agentic models. And when everything is digital, everything is a potential target.

• How will we use AI agents to monitor and disrupt other AI agents?
• What new standards for privacy and security will become necessary in the agentic economy?
• What happens to the security budget in enterprises as the pace of automated attacks accelerates?
• How will governments respond to accidents and attacks made possible by AI?

Reach out to Aditya

What if we could program the physical world as fluidly as we program software?

VLMs, world models, and rapid hardware iteration are collapsing the gap between bits and atoms, making physical systems more programmable, testable, and debuggable.

• What becomes possible with an API to the physical world?
• What does great “developer experience” look like for physical objects and spaces?
• Which physical problems become solvable with true spatial reasoning?

Reach out to Finn, Rui

How will AI inspire new telemetry: what we can perceive, measure, and understand about reality?

AI represents a new epistemic instrument—able to probe systems through embedded agents, not just observe—revealing patterns and structures humans cannot perceive.

• What telemetry emerges when agents act as embedded experimenters?
• How will scientific theories evolve under AI-generated causal insights?
• What becomes measurable with continuous, multi-scale instrumentation?
• What translation layers make AI-derived signals legible?
• How do we validate discoveries that may be incomprehensible to humans?

Reach out to Danh, Wei

How can AI enable deeper human connection rather than replace it?

AI often feels distancing, but it could instead serve as a catalyst for authentic relationships, surfacing shared interests, prompting timely touch points, bridging cultural gaps, and forming new kinds of communities.

• How can AI surface latent relationships and “you should meet” moments at scale??
• What new mediums or artifacts for connection become possible?
• How can AI genuinely bridge cultural and linguistic divides?
• How do we design for facilitation, not dependency?
• What new forms of community emerge when AI handles memory and logistics?

Reach out to Evan, Divanny

What new forms of governance will emerge from artificial intelligence and agentic economies?

AI introduces non-human actors and decision systems operating at speeds and scales that break existing structures. This may reshape sovereignty and accountability as we know it.

• What happens to the law when agents act autonomously?
• What will the role of the human be in the agent economy?
• What will change when agents have perfect memory and access to all information?
• ⁠What does accountability look like for businesses and individuals for the actions of their agents?
• ⁠How will lobbying and political influence transform when machines can advocate?

Reach out to Dylan, Sam

How should we rethink institutions to support rapid onshoring of manufacturing and production across sectors?

America is attempting the largest industrial reshoring in generations. We need new systems: credentialing that signals and verifies competence, career pathways that attract ambitious talent, and training models that keep pace with rapidly evolving manufacturing processes.

• How do we credibly train while also verifying that the work is high quality?
• What kinds of credentials or brands would make “advanced manufacturing alum” a status identity for ambitious people rather than a fallback for displaced workers?
• How would we design pathways in advanced manufacturing that feel more prestigious and upwardly mobile?
• Who pays for these institutions: workers betting on future careers, enterprises desperate for talent, or governments subsidizing strategic capacity?

Reach out to Arian, Danilo

If you're exploring any of these questions, feel free to reach out to the SPC team member listed, or apply to SPC.