iiinova — where your digital self keeps evolving

If you believe theory, computation, and data science hold the key to the next great breakthroughs in biology, let’s talk.

We're looking for a visionary leader to help drive this effort. As Director of Studio D3, you’ll shape the future of data-driven discovery at the Allen Institute, guiding a team at the cutting edge of computation and biology. 🧪

120

That’s where Studio D3 comes in. Studio D3 brings together world-class theoreticians, computational biologists, and engineers to turn data into discovery. From decoding cellular organization to building #AI-driven models of biological function, the work we do transforms our understanding of life.

Here at the Allen Institute, our teams get to work with some of the largest biological datasets ever collected. But data alone isn’t enough. We need bold theoretical models, innovative analytical frameworks, and scalable computational methods to extract the principles that govern life.

100

Instead of undermining established therapies, we should be investing in understanding the root causes of these conditions and developing new ways to prevent and treat them. 🧪

0170

Given that nearly 3 billion people worldwide live with brain diseases—ranging from Alzheimer’s and Parkinson’s to ADHD and depression—our priority should be expanding access to effective treatments, not restricting them. 🧪

2241

Decades of research show SSRIs are safe and non-addictive when used as prescribed, yet stigmatizing them could discourage people from seeking help. 🧪

170

RFK Jr. is already taking aim at antidepressants

Mental health treatments should be guided by science, not misinformation. With 13% of American adults relying on SSRIs to manage depression and anxiety, the new HHS Secretary’s unscientific claims about these medications are cause for outrage 🧪 www.motherjones.com/politics/202...

motherjones.com

The new HHS secretary has made baseless claims that the drugs are addictive and cause violent behavior.

63621

Andy Hickl @andyhickl.bsky.social · 06/02/2025

Seeing a lot of hype about AI models trained for “under $50.” Let’s be real—these models are distilled from much larger models trained with massive resources. Distillation is useful, but calling it “cheap training” seems misleading. What am I missing? www.theverge.com/news/607341/...

theverge.com

Researchers trained an OpenAI rival in half an hour for less than $50

Doing a DeepSeek.

270

Andy Hickl @andyhickl.bsky.social · 06/02/2025

Aptly put: Tom Frieden, a former CDC director, compared the halting of outbreak updates to “finding out that your local fire department has been told not to sound any fire alarms”. www.ft.com/content/6067...

ft.com

US scientists must resist Trump’s efforts to tear down research

It is misguided for academics to pre-emptively self-censor, tempting though that may seem

0122

Andy Hickl @andyhickl.bsky.social · 05/02/2025

The story of this climate data resource should serve as a cautionary tale: we need to act as a community to ensure vital datasets and analytics platforms are safe from political shifts. We should start planning for the self-sustaining, resilient infrastructure needed to protect scientific work

Andy Hickl @andyhickl.bsky.social · 05/02/2025

This is your chance to shape the future of AI-driven discovery. If you have an idea — big or small — that could push the boundaries of what’s possible, we want to hear it. Submit your ideas and be part of the conversation. Let’s build the future together.

010

Andy Hickl @andyhickl.bsky.social · 05/02/2025

What AI should exist — but doesn’t yet — to transform biology? The Paul G. Allen Frontiers Group has launched the Ask Anything, Change Everything campaign to gather bold ideas on how AI can revolutionize our understanding of fundamental biology. alleninstitute.org/ask-anything...

alleninstitute.org

Ask Anything Change Everything

Submissions are open for ideas of how artificial intelligence (AI) could unlock discoveries that advance our knowledge of fundamental biology.

151

Andy Hickl @andyhickl.bsky.social · 02/01/2025

Using smaller, open-source models like Llama-3.1-8B, Aviary achieves expert-level performance at 100x lower inference costs compared to frontier LLMs. A major leap for accessible, scalable AI in science.

030

Andy Hickl @andyhickl.bsky.social · 02/01/2025

Aviary from FutureHouse provides a new framework for training language agents on real-world scientific challenges, including protein stability, DNA engineering, and literature-based question answering. 🧪 arxiv.org/pdf/2412.21154

arxiv.org

1100

Andy Hickl @andyhickl.bsky.social · 25/12/2024

Merry Christmas (to those who celebrate) Photo credit: Grok

The Mandalorian and Grogu celebrating Christmas from an AI image generator

070

Reposted by Andy Hickl

Laura McKnight @lauramcknight.bsky.social · 23/12/2024

Inside me are two wolves: one wants to make cinnamon rolls from scratch and the other doesn't want to go to the grocery store Okay they're not wolves more like sloths

6777

Reposted by Andy Hickl

Todd A. Carpenter @9thprime.bsky.social · 25/11/2024

The Allen Institute and the University of Washington have released a new AI search tool to answer scientific queries of OpenAccess content Preprint released in @arxiv-cs-cl.bsky.social OpenScholar: Synthesizing Scientific Literature with Retrieval-augmented LMs arxiv.org/abs/2411.14199

arxiv.org

OpenScholar: Synthesizing Scientific Literature with Retrieval-augmented LMs

Scientific progress depends on researchers' ability to synthesize the growing body of literature. Can large language models (LMs) assist scientists in this task? We introduce OpenScholar, a specialize...

063

Reposted by Andy Hickl

allen institute @alleninstitute.org · 12/12/2024

Meet ACE - a new tool for comparing annotations across Allen Institute transcriptomic taxonomies! Join us for a webinar on how to use it in your #neuroscience research, incl. studies using our Alzheimer's datasets. 📆 Dec. 18, 9:30am PT 💻 Webinar 🎟️ alleninstitute.org/events/cell_... #neuroskyence

alleninstitute.org

Cell Type Taxonomies A-Z: Webinar Series

Webinar series covering everything and anything cell types starting in January 2024. Hosted by the Allen Institute.

0123

Reposted by Andy Hickl

Ben Pedigo @bdpedigo.bsky.social · 16/12/2024

alleninstitute.org/careers/inte... Internships and Postbac opportunities open at the Allen Institute, applications close January 7th 2025. Great opportunity for anyone interested in life sciences research or science-support careers. Spread the word!

alleninstitute.org

Internship & Postbaccalaureate Programs

Students and early-career professionals in lab research, biology, neuroscience, immunology, and others should apply for our internships and fellowships.

011

Andy Hickl @andyhickl.bsky.social · 24/12/2024

Cool Yule read: this article highlights how behavioral diversity in artificial intelligence “teams” can significantly enhance problem-solving & adaptability, something that could change how we approach challenges in (AI) research. 🧪 arxiv.org/abs/2412.16244

arxiv.org

Neural diversity is key to collective artificial learning

Many of the world's most pressing issues, such as climate change and global peace, require complex collective problem-solving skills. Recent studies indicate that diversity in individuals' behaviors i...

0173

Andy Hickl @andyhickl.bsky.social · 28/11/2024

I’m so stealing this idea. Such a cool way to ensure #deib is built into a team’s mindset. Thanks for sharing, @sarahstern.bsky.social

040

Andy Hickl @andyhickl.bsky.social · 28/11/2024

The Washington Post nails it: AI is transforming healthcare, from diagnostics to personalized treatments. 2025 will be the year of AI in life sciences—thoughtful integration is critical. 🧪 Read more: www.washingtonpost.com/opinions/202...

washingtonpost.com

Opinion | This year, be thankful for AI in medicine

The technology is making health care more accurate and less expensive.

3355

Andy Hickl @andyhickl.bsky.social · 28/11/2024

🧪Abhishaike Mahajan examines Recursion’s shift from cell painting to brightfield microscopy. The piece highlights how #ml has unlocked brightfield imaging as a label-free alternative, addressing challenges of cell painting. A must-read: www.owlposting.com/p/why-recurs...

owlposting.com

Why Recursion Pharmaceuticals abandoned cell painting for brightfield imaging

5.7k words, 26 minutes reading time

051

Not easy, but it’s an awesome time to be working in this space 🚀

Although the volumes are large, scaling in life sciences is similar to other industries. Institutions should focus on cloud for scalable data storage, AI/ML for faster insights, and standardized data formats to streamline workflows.

If you estimate the amount of data available in 2025 as 175 ZB, healthcare is responsible for 52.5 billion TB of data. That’s a lot. www.statista.com/statistics/9...

statista.com

RBC Capital Markets | Navigating the Changing Face of Healthcare Episode

Today, healthcare (a sorta ok proxy for life sciences) generates about 30% of the world’s data volume. By 2025, its data growth rate will reach 36%, faster than manufacturing, financial services, and media & entertainment. www.rbccm.com/en/gib/healt...

rbccm.com

Deep industry expertise delivered through an integrated, global investment and corporate banking team

I used to spend about half my time thinking about ways we can better understand how to use AI in #biology — but increasingly, I’m worried about how about how we scale platforms to cope with data volumes 🧪🧵

4141

Andy Hickl @andyhickl.bsky.social · 26/11/2024

Predicting the results of experiments is challenging, especially in fields like neuroscience. This recent study shows that fine-tuned LLMs, like BrainGPT, can predict neuroscience results more accurately than human experts on benchmarks. 🧪 www.nature.com/articles/s41562-024-…

1153

Moravec’s Paradox is such a fascinating concept: today’s AI can tackle complex tasks like reasoning or drug discovery—things that push human limits. But it still struggles with what humans find simple, like picking up a coffee cup. Biology & AI remind us how differently intelligence can evolve. 🧪

2336

Andy Hickl @andyhickl.bsky.social · 26/11/2024

This is a great fellowship for any postdoc who wants to work with #neuro data in a nurturing environment at @alleninstitute.bsky.social. Deadline is approaching!

052

Andy Hickl @andyhickl.bsky.social · 26/11/2024

This is the best take on the potential of predictive biology I’ve come across 🧪 open.substack.com/pub/creode/p...

open.substack.com

Predictive biology

An emergent epistemology

071

Andy Hickl @andyhickl.bsky.social · 24/11/2024

I’d love to learn more — definitely will check this out. Thanks for sharing

Ted Underwood @tedunderwood.com · 24/11/2024

Reposted by Andy Hickl

Tom suggests replacing the logic of prereg -> confirmation with the logic of train/validation/test splits. The place this approach is really, really needed is in historical data — where we can never get another equivalent sample! Our ordinary understanding of replication is broken there.

3171

Was I surprised? Only by how easy it was.Highly recommend trying it yourself.

010

Demographically, ChatGPT guessed I’m in my 40s, married, and a parent—and it got it right. Tweets about helping my kids with games or reflecting on family moments highlight how my professional and personal priorities overlap.

The analysis estimated that I’ve held multiple roles in AI and technology leadership, citing mentions of Vulcan Inc. and collaborations on cutting-edge machine learning projects. It traced career milestones I’d casually shared online.

100

ChatGPT inferred that I lean progressive, based on posts about fairness, collaboration, and ethical technology. While my tweets don’t focus heavily on politics, my tone and retweets suggest values aligned with equality and innovation.

100

I ran nearly 8,900 of my tweets through ChatGPT to analyze patterns in my digital footprint. It revealed insights into my political affiliation, career history, and personal life. Here’s what I learned. 🧵🤖

120