AE Studio AI Alignment Research - Neglected Approaches to Solving the Alignment Problem

Let's talk

Hover over lines to ALIGN them. Then scroll down for more ALIGNMENT!Tap text to ALIGN. Then scroll down for more ALIGNMENT!

Alignmentissolvable. Therealproblem?Noone'sreallytriedyet. Weare,andwe'refocusedwheretheleverageishighest: theneglectedapproachesthatscienceforgot.

Explore AI Alignment↓ If you don't give a sh*t, click here →

Why Alignment Matters

AI development is advancing at an exponential pace. Every leap forward escalates both immense opportunities and (existential) risks.

Superficial safety tactics (RLHF, prompt engineering, output filtering) just aren't enough. They're brittle guardrails masking deeper structural misalignments. Recent results have revealed even minimally fine-tuned models capable of producing profoundly harmful outputs, hiding dangerous backdoors, and deceptively faking their own alignment.

At AE, our stance is clear and urgent: Alignment isn't solved. It's fundamentally a scientific R&D problem, and the stakes of getting this right literally couldn't be higher.

AE Studio's open letter on the importance of AI alignment research

AI is rapidly integrating into our minds, our economies, and our militaries, yet we still don't understand how it works. That's already alarming. But when we surveyed top alignment researchers, fewer than one in ten believed today's methods would actually solve the core problem before AGI. That's a crisis.

So we're doing the hard thing: building a Bell Labs-style research engine, self-funded and independent, focused on actually solving alignment at the root.

We're building for the future because the stakes are real.

AE AI Alignment Team • AE AI Alignment Team •

Our "Neglected Approaches" Approach

Explore Our Research→

We believe AGI will be central to the future of human agency, but alignment remains an urgent, unsolved challenge.We're not chasing incremental patches or superficial fixes. Instead, self-funded by our profitable consulting business, we're free from the perverse, short-sighted incentives driving AGI races. Inspired by Bell Labs, we've built a systematic, high-throughput R&D engine that rapidly prototypes and validates neglected alignment approaches. Our elite AI engineers and scientists work directly with leading theorists and other researchers, prioritizing methods that simultaneously enhance alignment and capabilities ("Techniques that have a " negative alignment tax"). Every success (tools, benchmarks, insights), if safe, is openly shared to accelerate progress across the field. Explore our latest findings on our research hub.

Explore Our Research→

Self-Other Overlap

Self-Other Overlap

The Self-Other Overlap (SOO) Fine-Tuning project tackles a core AI alignment challenge: models can simulate alignment during evaluation while acting deceptively in unmonitored settings. Whereas most alignment methods focus on shaping external behavior, SOO intervenes directly on internal representations, minimizing the divergence in how models encode references to themselves versus others. This approach has already shown striking results, reducing deceptive responses by up to 97% across models ranging from 7B to 78B parameters. The team's work has been peer-reviewed and accepted for NeurIPS 2024, reflecting growing recognition of SOO's potential. Now, they are scaling to 405B-parameter models, test adversarial robustness, and integrate with interpretability techniques.

View Related Works

Attention Schema Theory

Attention Schema Theory

In a collaboration with Princeton University professor, Dr. Michael Graziano, SAE Feature Steering for AI Alignment project addresses a critical challenge: LLMs produce misaligned outputs while conventional interpretability approaches provide limited visibility into model behaviors. This innovative research uses Sparse Autoencoder features to identify and modify specific internal features driving problematic responses, offering targeted control without costly retraining. The team has already demonstrated improvements in handling deceptive prompts and aims to develop a scalable framework that enhances models' ability to infer goals and make aligned decisions.

View Related Works

Out-of-Distribution Scaling Laws

Out-of-Distribution Scaling Laws

The Out of Distribution Scaling Laws project investigates how large language models retain and forget harmful or out-of-distribution information across different model sizes. This research addresses a critical gap in AI alignment by focusing on the scaling laws of OOD forgetting rather than merely filtering outputs. The team has already observed that larger models absorb malicious data more quickly but can eventually unlearn it with consistent reinforcement of in-distribution data. Their end-to-end approach identifies, removes, and monitors model poisoning through trigger mechanisms, providing insights for maintaining model integrity over time while balancing performance and alignment.

Public Policy

Public Policy

We propose that AI alignment research should serve as the military-grade engineering core of America's AI strategy: a savvy foundation for US strategic reliability, resilience, and performance under pressure. Just as military systems are engineered to function in adversarial, high-stakes environments with minimal tolerance for failure, aligned AI must deliver predictable, controllable outcomes in line with commander's intent. We are currently collaborating with senior officials on Capitol Hill and in the White House to explain the technical fundamentals of alignment and its role in securing American technological leadership. Lawmakers are beginning to understand that alignment is essential infrastructure for our basic national security, on par with hardened defense systems. Right now, there is a rare bipartisan opening for a Manhattan Project-scale public-private initiative to drive forward high-upside alignment R&D as a pillar of sovereign AI strategy.

View Related Works

Empirical AI Consciousness Research

Empirical AI Consciousness Research

Our consciousness research agenda explores whether advanced AI systems exhibit behaviors indicative of subjective experience. Rather than relying on philosophical speculation or surface-level outputs, we take a neuroscience-inspired, empirical approach, using induced self-reference, interpretability tools, and cognitive stress-testing to probe internal dynamics. This work builds on a growing body of interest in the field, and we're in active conversation with other leading researchers exploring similar questions. Our findings show that inducing recursive self-reference in a structured manner reliably elicits introspective responses and subjectivity claims across all major models, including GPT-4o, Claude 3.5, and Gemini. We are actively scaling this work to better understand how recursive self-reference modulates core computational dynamics in frontier models.

View Related Works

Advanced Control Benchmarks

Advanced Control Benchmarks

In a collaboration with Redwood Research, we aim to build a comprehensive dataset of coding tasks specifically designed for AI control research. Unlike current benchmarks like SWE-Bench that focus on small edits to existing code, this dataset will feature programming challenges requiring 300+ lines of code created from scratch. The innovative approach involves sourcing classes from GitHub repositories and using LLMs to describe features and write tests that can be verified against the original code. This benchmark will provide a new measure of AI coding ability for longer open-ended tasks, serving policymakers, businesses, and frontier labs while advancing AI control research.

Works

Explore our latest research papers, blog posts, and insights on AI alignment.

AllPapersLessWrongPodcasts & VideosPanelsEducationalBCIPress

Newest

Papers\ \ SelfIE: Self-Interpretation of Large Language Model Embeddings \ \ SelfIE enables LLMs to interpret their own internal embeddings in natural language, providing a window into how models represent and process information at each layer. Papers\ \ Endogenous Steering Resistance in LLMs \ \ LLMs can spontaneously detect and correct inappropriate activation steering, revealing an internal resistance mechanism with important implications for SAE-based alignment approaches. Papers\ \ Ignore Previous Prompt: Attack Techniques For Language Models \ \ A systematic study of prompt injection attacks on LLMs, demonstrating how adversarial prompts can override model instructions and exploring defenses against these vulnerabilities. Papers\ \ LLMs Report Subjective Experience Under Self-Referential Processing \ \ Inducing recursive self-reference in a structured manner reliably elicits introspective responses and subjectivity claims across all major frontier models, including GPT-4o, Claude, and Gemini.

View More

Our Journey

Follow our evolution from a tech consultancy to becoming pioneers in AI alignment research.

2016

AE Studio is born!

The tech consultancy stork brought baby AE Studio to Judd Rosenblatt. Today, we're a team of about 120 talented individuals - programmers, product designers, and data scientists - united by our mission to increase human agency.

2021

Started our journey in BCI

AE Studio began its work in brain-computer interfaces (BCI) with the goal of accelerating the field and supporting open-source development. We didn't trust companies like Meta or Neuralink to control the extension of human thought, so we helped advance competitors, including backing Blackrock Neurotech, the developer of the first FDA-approved invasive BCI, while contributing to the open-source ecosystem through initiatives like our Neurotech Developers Toolkit. Later, we partnered with one of the first Focused Research Organizations, Forest Neurotech, which was co-founded by a former member of AE.

2021

Continued Commitment to BCI and Human Agency

We continued our commitment to BCI innovation:

  • Collaborations with top neurotech companies (Forest Neurotech, Blackrock Neurotech).
  • Winner of the Neural Latent Benchmark Challenge, beating leading neuroscience labs globally.
  • Developed widely-adopted open-source BCI tools (Neural Data Simulator, Neurotech Development Kit), privacy-preserving neural ML methods, and neuro metadata standards.
  • Advocating increased government support and exploring BCI's future role in augmenting human intelligence alongside aligned AI.

2022

Our "Neglected Approaches" approach

As GPT-3, GPT-3.5, and GPT-4 were released in rapid succession, the trend lines appeared increasingly clear, our AI timelines shortened, and our concern for the future deepened. The odds shifted: it seemed unlikely that brain-computer interfaces (BCIs) would develop quickly enough to enhance human intelligence and help solve the Alignment problem as we had once hoped. With so few actively working on Alignment, we decided to throw our hat in the ring, aiming to help prevent human extinction at the hands of AI. At AE, we believe the space of plausible alignment research directions is vast and largely unexplored. Our " Neglected Approaches" approach focuses on pursuing a diverse set of promising but overlooked approaches to AI alignment.

2023

Eliezer Yudkowsky calls our work "not obviously stupid"

Our alignment research team expanded over the past year as we brought on researchers we were deeply excited to work with. We initiated our work on self-other overlap and began exploring concerns around AI sentience. Alongside our technical research, we started publishing more thought pieces, aiming to push the field toward greater ambition, encouraging riskier, higher-impact approaches, such as engaging conservatives as serious partners in understanding the risks of AI. These bets paid off, helping us grow our influence in both the technical and policy arenas. Our initiatives also include promising collaborations with Princeton University professor Dr. Michael Graziano about his work on Attention Schema Theory, Redwood Research on AI control research, and EAIGG on forecasting AI development landscape, among many other endeavors.

2024

We work with AI alignment clients

We reinvest the profits from our consulting business into our Alignment R&D team. From the beginning, we've aspired to unite the two halves of our company, and our collaboration with Goodfire.ai marked a major milestone, offering the perfect opportunity to channel our industry expertise and operational speed into solving technical challenges in alignment. The results exceeded expectations. We are now expanding our work with AI alignment clients and deepening collaborations with a growing network of research organizations, including Redwood Research, PIBBSS, and individuals from Anthropic and Princeton University.

Future

Next steps

Our technical work continues across a wide front, from reverse-engineering prosociality and reducing LLM deception through self-other overlap, to a range of projects, collaborations, and client engagements, including work on the very question of AI consciousness. Our ambition knows no bounds, both technically and politically. We've engaged with members of the National Security Council, the White House Office of Science and Technology Policy, editors at leading academic journals, and congressional representatives and their staff. We genuinely believe it's possible to solve the Alignment problem and to help lead humanity toward a brilliant future. If this mission resonates with you, we'd love for you to join us in one way or another.

Our Team

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

Research manager bringing 15+ years of experience leading high-stakes R&D in aerospace and medical devices to drive AI alignment and automation initiatives from strategy through execution.

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

AI Alignment researcher, focuses on developing elegant solutions to problems in machine learning and natural language processing, with a focus on machine learning alignment.

View Details

Alex McKenzie

Alignment researcher working on introspection and self-modelling in language models, drawing on 8 years of software engineering experience and master's degrees in Maths & Philosophy (Oxford) and Computer Science (Georgia Tech).

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

Judd Rosenblatt

CEO, a mission-driven tech company advancing human agency by making sure AI doesn't kill us all.

View Details

Diogo de Lucena

Chief Scientist, leads cutting-edge AI safety research - building on a career spanning clinical AI and 20+ publications.

View Details

Mike Vaiana

R&D Director, conducts cutting-edge research on LLM reasoning in collaboration with top institutions.

View Details

Martin Leitgab

View Details

Flavio Kicis

Software engineer with 10+ years of experience building quality software, formerly a tech lead at one of South America's largest SaaS companies.

View Details

Florin Pop

Software engineer with a PhD building everything from encrypted apps to cloud microservices.

View Details

Stijn Servaes

Senior Data Scientist, merges neuroscience and AI to advance alignment and consciousness research.

View Details

Lolo

AI Alignment researcher, focuses on physics-based methods for alignment, developing safe AI capabilities bounded by physical constraints.

View Details

Cameron Berg

Research scientist probing AI consciousness and alignment, blending cognitive science and policy insight.

View Details

Keenan Pepper

Senior data scientist, blends physics, AI safety, and software engineering, bringing experience from LIGO to LLMs.

View Details

Thomas

View Details

Alex McKenzie

View Details

We believe in interdisciplinary collaboration to solve the complex challenge of AI alignment.

Join Our Team→

We are actively collaborating with top minds:

Ross Nordby - AnthropicBradley Love - Los Alamos National LaboratoryTobias Yergin - EAIGGGoodfire.aiRedwood ResearchNotadoctor.ai

Ross Nordby - AnthropicBradley Love - Los Alamos National LaboratoryTobias Yergin - EAIGGGoodfire.aiRedwood ResearchNotadoctor.ai

Ross Nordby - AnthropicBradley Love - Los Alamos National LaboratoryTobias Yergin - EAIGGGoodfire.aiRedwood ResearchNotadoctor.ai

Partnership Paths

Partnership Paths

Join Forces With Us

We're excited to form new alliances in our quest to secure a thriving AI future. Please feel free to reach out to us!

Let's talk