Sapien Weekly Digest

Hey there, Sapiens!
Happy Halloween! Tonight is the night of scares and pranks, but first, let’s look at what has happened on Sapien’s end this week:

What we’ve been up to

Hallo-Meme Contest.
The first Sapien Hallo-Meme contest has drawn to a close, and we’re proud to present the winners! Our team has voted internally to choose the memes that made us laugh the most.

From user franklovesNFT on Discord, here is our third place:

Moving on, here’s our second place finish, courtesy of user imakhanhdns on Discord:

And finally, our grand prize of 2,000 SAPIEN will go to user elonman on Discord thanks to the following meme:

I want to give a massive shoutout to all the Sapiens who competed, some of you even submitted multiple memes that made us laugh! Please look out for the next community event as there’s more to come!

New Hires.
Two new faces have joined the Sapien team. We’re happy to welcome Adam Li, our new Data Scientist, and Ali Malik, our new Product Manager.

Adam joins us from British Columbia in Canada, bringing deep academic and applied machine learning experience from UC San Diego and Harvard Medical School to CBC and Kero Sports. He’s passionate about building systems that bridge scientific rigor with real-world impact.

Ali, based in Ontario in Canada, brings his product craft from YC startups like Surf and Tempo, where he turned fast-moving ideas into beautifully functional user experiences, studying at the London School of Economics while working full time.

Current task Overview.

🍳 Breakfast Image Upload
🍽 Dinner Image Upload
🧩 Logic Path – Easy / Intermediate / Hard
🌐 Cadena Lógica (Español)
🗣 Njia ya Mantiki (Kiswahili)
🧠 Đường dẫn lô gic (Tiếng Việt)
🧬 Gastro Tag
🐞 Bug Hunters
📦 Breakfast Image Bounding Boxes

Since last week, we’ve updated to Sapien v0.185.0 - 

🚀 What’s New

• Streak multipliers

• Demographic restrictions for tasks

• UX fixes and improvements

• Synchronization of fonts across app

• Bug fixes related to staking and rewards

• New project preparation - car model matching

Our Voices in the World.

“AI actually needs humans more than humans need AI.”
NBC has featured us in a short clip on how AI is affecting the job market, watch our co-founder Trevor Koverko’s take on why human intelligence remains essential.

Our research this week focused on what happens when AI runs out of freely available data on the internet to train on. We’re hitting this roadblock as the sheer size of datasets used for AI models is rapidly exhausting the energy grid. The solution, as so often in this space, is to make sure that there is a continuous human in the loop to keep AI accountable.

This week, we were invited as guests on a space with our good friends at W3 talking about Artificial Intelligence in Healthcare and the hurdles we still have to overcome before it can be commonplace.

What else is new?

The first major settlement in AI Music.
UMG and Udio have settled a copyright‐infringement lawsuit that UMG, along with other major labels, brought against Udio. As part of the settlement, Udio will now operate under a licensing framework for UMG’s catalog of both recordings and publishing when using it for its AI music generation platform. UMG artists and songwriters will be compensated for use of their work in both training of AI and in AI‐generated outputs. This is one of the first major settlements of its kind between a major label and an AI music‐generation platform, moving the industry from litigation toward licensing frameworks.

Google Earth now runs on AI.
Google has launched Google Earth AI, a platform that fuses satellite imagery with powerful AI models to help address environmental crises like floods and wildfires.
Google describes it as a collection of geospatial AI models and reasoning agents designed to provide actionable insights about the planet, grounded in satellite imagery, other Earth observation data, population and environment models. The tool aims to enable governments, researchers, and disaster-response teams to detect risks, track changes, and respond faster.

GPT-5 as a therapist.
OpenAI just rolled out a major update to GPT-5, designed to better recognize and respond to users experiencing mental health crises. After consulting with over 170 mental health experts across the globe, the AI now flags signs of distress more reliably, and routes users to more capable models and crisis resources. The company reports that in these domains the new model reduced the rate of responses that “do not fully comply with desired behaviour” by about 65‑80% across production traffic for mental‑health‑related areas. OpenAI estimates over a million users weekly show signs of severe distress, prompting a push toward more responsible AI intervention.

Where are we going

With the Hallo-Meme contest being as much of a success as it was, expect more ways to interact with the community moving forward!

In the Pipeline:

Early Onboarding. We’re continuing to test ways for a smoother onboarding experience for new users, including explainer videos and more ways to get to your first tasks.

Better matching. We’re working on systems that will identify you as a strong fit for a given task rather than waiting for you to find it in the app.

Consistency incentives. We’re aiming to make it easier and more rewarding to be consistent.

Remember to invite a friend! Sapien thrives on human insight. Your ideas, humor, and curiosity drive everything we build. Every new Sapien strengthens our collective intelligence. Share the app and grow the movement.

Join the Sapien Discord for more events to come!