Sapien Weekly Digest

Hey there, Sapiens! Welcome to the last Weekly Digest of the year! 

What we’ve been up to

Sapien Builder Casting Call:
We’re looking to identify the builders in our community who are already working with agents, LangChain/LangSmith, and onchain automations. 

If you’re:
- Building or experimenting with AI agents  
- Using LangChain, LangSmith, or tools like Coinbase AgentKit  
- Running any kind of on-chain or autonomous transaction flows  
…we really want to hear from you.   

Your answers directly influence our roadmap and how we support builders going into 2026. Fill out the form here!

Sapien Team Takeover:
Today, our CEO Rowan Stone mingled with the community and chatted with you all about the future plans for 2026 and what we have in store for the community. He went into detail about our Proof of Quality strategy for the new year as well as the first mover advantage it brings with it. He also addressed task availability and what our plans are to make sure the community has plenty to do! Read the full AMA here!

🚀 What’s New in the App?

We’re sending you into the holidays with the following tasks to work on, with one more planned to be added!

Current task Overview:

1. 👩‍🍳 Gastro Tag – Easy / Intermediate / Hard

2. 👩‍🍳 Gastro Tag QA – Easy / Intermediate / Hard

3. 💬 Emotion Prompt – Easy / Intermediate / Hard

4. 💬 Emotion Prompt QA – Easy / Intermediate / Hard

5. 🧩 Logic Path Vietnamese QA

6. 👋 Hand Object Annotation

7. 📦 Breakfast Image Bounding Boxes

8. 🚗 Vehicle Model Mapping

New Tasks:

Multi-Choice Assistant Error Review is coming to the app very shortly, as well as a Quality Assurance variant! More to come about what this task will look like in detail, but we will have to be patient just a little while longer on this one.

Our Voices in the World

“The most important part is setting the proper guardrails. As we project what these agentic ecosystems could become, we need to ensure we can trust that their autonomy isn’t heading in the wrong direction.”

This Wednesday, our Director of Solutions Architecture, Lukas, was invited by our friends at W3 to speak about AI Agents and how to build an economy around them. He went into detail about agentic economy fundamentals and how to properly manage agent complexity. Find the full recording right here!

“We are increasingly bringing AI from the digital world to the physical world, and it’s critically important that we do that in a very careful and considerate way,” 

Our CEO Rowan Stone was quoted by TheStreet in their coverage on Waymo and the public perception of autonomous vehicles as part of an attempt to highlight what can be done to ensure safety in edge cases. Read the full article here!

Where are we going?

A detailed Roadmap! In the new year, we’re planning on going big! More partnerships, more events, and more opportunities to build on Proof of Quality.

What else happened in AI?

The World outside of Sapien:

U.S. Department of Energy expands the Genesis Mission:
On December 18, 2025, the U.S. Department of Energy announced memorandums of understanding with 24 organizations to support the Trump administration’s Genesis Mission, a White House directed effort to apply AI and advanced computing to accelerate scientific discovery across the DOE national laboratory system. The White House executive order launching Genesis Mission frames the effort as comparable in urgency and ambition to the Manhattan Project, stating the project will focus on federal scientific datasets, scientific foundation models, and AI agents to automate research workflows.
To that end, DOE will mobilize its 17 National Laboratories, containing roughly 40,000 scientists and engineers, aiming to double the productivity and impact of American research within a decade.  DOE has two open RFIs tied to Genesis Mission partnership activity, with deadlines on January 14, 2026 and January 23, 2026, which will indicate where DOE wants additional partners and what work packages are imminent.

Google Releases Gemini 3 Flash:
After OpenAI released GPT Image 1.5 earlier this week, Google hit back with Gemini 3 Flash on December 17, 2025 and made it the default model in the Gemini app and Google Search AI Mode, positioning it as a low latency, lower cost model that approaches flagship level quality while materially improving unit economics. Google says Flash delivers frontier class results like 33.7% on Humanity’s Last Exam without tools and lists API pricing at $0.50 per million input tokens and $3 per million output tokens, undercutting Gemini 3 Pro pricing by four times at standard rates.

The AI Google search function now defaults to Flash and continues to provide real time information and links from across the web

Stanford University predicts a major 2026 for AI:
Stanford University’s Institute for Human-Centered AI forecasts 2026 as a reset year for artificial intelligence, moving from hype to measurable impact. Experts there predict no breakthrough toward artificial general intelligence (AGI) in 2026, and expect more firms to acknowledge that AI has not delivered broad productivity gains outside areas like coding and call centers. In healthcare, researchers anticipate a “ChatGPT moment” as training costs for medical models fall and access to clinical datasets rises, enabling more capable, clinically useful systems. In the legal sector, firms will shift from asking if AI can write to evaluating how well it performs, on what tasks, and with what risks, especially for complex, multi-document reasoning. The collective view is that 2025’s hype and massive investment will give way to 2026’s demand for evidence of real value and risk-adjusted ROI. Stanford HAI does not predict an AI bubble crash but expects patience for overpromises and pilot demos to erode. 

HAPPY HOLIDAYS to all Sapiens! See you in the new year!