Can neuroengineering address AI alignment or the Intelligence Curse?

commitment: 10 hrs/week
flexibility: Open to proposals

mentored by

The project

A growing but fragmented set of proposals suggests that neuroengineering could help humanity navigate the development of advanced AI. These include whole brain emulation, brain computer interfaces, and neuromorphic or brain inspired AI. However, these technologies are often proposed as responses to two distinct problems.
The first is AI alignment, meaning how to ensure that advanced AI systems remain compatible with human values and subject to meaningful human direction. The second is the Intelligence Curse, the possibility that increasingly capable AI could reduce the economic and political value of human labour, expertise, and participation, gradually weakening human agency and influence.
Several researchers have proposed different neuroengineering responses. Michael Florea argues that biological intelligence may need to develop quickly enough to remain competitive with artificial intelligence. Tom Burns has explored whether brain computer interfaces could improve human oversight and intervention, while Wei Dai has proposed that idealized whole brain emulations might direct advanced AI systems. Research on brain inspired AI has also examined whether mechanisms associated with human motivation, empathy, or social cognition could contribute to safer and more prosocial systems.
These proposals do not all make the same claim. Some suggest that neuroengineering could produce intelligence that is easier to align. Others focus on improving human oversight or increasing the capabilities of human decision makers. A third group suggests that augmentation could help preserve human relevance and extend the period during which human and AI teams remain competitive with fully autonomous systems.
There are also important reasons for skepticism. Reproducing features of human cognition would not automatically create aligned systems, and whole brain emulation may not arrive on useful timelines or remain competitive with purpose built AI. Neuroengineering could also accelerate capabilities while introducing new risks involving consent, identity, inequality, concentration of power, moral status, and the welfare of digital minds.
The central research question is:
How strong is the case that whole brain emulation, brain computer interfaces, or neuromorphic AI could address either AI alignment or the Intelligence Curse, and what new welfare and governance problems would each approach create?
The project would evaluate the two problems separately. For alignment, it would examine what problem each technology claims to solve, the mechanism through which it is expected to improve safety, the evidence supporting that mechanism, and whether it genuinely addresses alignment or merely relocates the problem.
For the Intelligence Curse, it would assess whether each technology could help humans remain economically, politically, or strategically relevant, what capabilities it would improve, whether it could scale on relevant timelines, who would control and access it, and whether it would preserve durable human agency or simply postpone displacement.
The expected output would be a structured evidence map and comparative scorecard assessing the strength of the arguments, available evidence, timelines, assumptions, and governance and welfare risks associated with each approach.

The mentee's role

Mentees would play a leading role in shaping and conducting the research. Together, we would refine the search strategy, define the coding framework, and determine whether the project should cover all three technologies or investigate one in greater depth.
Mentee tasks could include:

Conducting the structured literature search
Building and maintaining the evidence database
Coding sources according to the shared framework
Reconstructing the strongest argument for each proposal
Assessing evidence, assumptions, timelines, and counterarguments
Developing the comparative scorecard
Identifying governance and welfare challenges
Drafting and revising the research memo
Translating the findings into a public article or another accessible resource

Mentees would be expected to take ownership of the research between meetings while receiving regular guidance, detailed feedback, and support with project direction.

Who I'm looking for

I am looking for a mentee who is open minded, intellectually curious, and genuinely interested in exploring difficult interdisciplinary questions. The strongest fit will be someone who is comfortable working in an emerging area where the literature is fragmented, the evidence is uncertain, and reasonable people may disagree about the conclusions. Expected commitment
10 hours per week.

Must haves:

Intellectual curiosity and a willingness to engage seriously with unfamiliar ideas.
An open mind, including a willingness to challenge initial assumptions and update in response to evidence.
Strong reasoning and clear written communication.
Reliability, initiative, and the ability to make steady progress independently between meetings.
Comfort working across disciplines and distinguishing empirical evidence from theoretical argument or speculation.
Receptiveness to feedback and willingness to revise work iteratively.
The ability to commit approximately 10 hours per week throughout the program.
Genuine motivation to produce a concrete and useful output, such as an evidence map, comparative scorecard, research memo, or public article.

Nice to haves:

Basic familiarity with AI alignment, AI governance, digital minds, or the post AGI transition.
A background in neuroscience, cognitive science, computer science, philosophy, economics, public policy, or another relevant field.
Experience conducting literature reviews, evidence mapping, academic research, policy research, or analytical writing.
Familiarity with whole brain emulation, brain computer interfaces, neuromorphic AI, cognitive enhancement, or NeuroAI.
An interest in the governance and welfare implications of emerging technologies, including questions about consent, identity, moral status, artificial suffering, cognitive liberty, and the concentration of power.
Experience comparing competing arguments, evaluating the strength of evidence, or developing structured assessment frameworks.
Interest in translating complex research into clear and accessible public facing work.

Deep expertise in neuroscience or AI safety is not required. I care more about how someone thinks, how they approach uncertainty, and whether they are motivated to learn.

Questions for applicants

Feel free to answer either one:

Of the neurotechnology adjacent proposals discussed above, which do you think is most promising for helping humanity navigate advanced AI, and why? How far could it realistically progress over the next five to ten years, and what do you see as the main technical or governance bottlenecks?
One question I'm interested in is whether increasingly capable AI could gradually reduce humanity's agency, influence, or relevance, sometimes referred to as the "Intelligence Curse." What is your initial reaction to this idea? Which questions do you think are most important to investigate, and why?

Support offered

I enjoy helping mentees turn broad, ambitious questions into focused, decision relevant research. Early in the project, we'll work together to identify a tractable research question, develop an analytical framework, and scope the project into something that can produce a meaningful output within the incubator.
Throughout the program, I will provide strategic guidance, detailed feedback on outlines and drafts, accountability through regular check ins, and help navigating interdisciplinary literature spanning neuroscience, AI, governance, and philosophy. I enjoy discussing ideas, challenging assumptions, and helping researchers sharpen both their arguments and their writing.
I'm also a very responsive mentor. If you're stuck between our scheduled meetings, I'm happy to answer questions over Slack or email, schedule ad hoc calls when useful, review work as it develops, and make introductions where I think they could help. I genuinely want my mentees to succeed, and I'll do everything I reasonably can to support them throughout the program.
While I'll be highly available for guidance and feedback, I also expect mentees to take ownership of their projects. My goal is to help you become a more independent researcher by providing strategic direction, constructive feedback, and encouragement while you drive the work forward.

format: Research & writing · Product & entrepreneurship · Field-building & outreach · Project scoping
topic: Longtermism · Post-AGI transition · Other · AI safety & alignment

Hyacinth Zia

Iliad

I have a background in neuroscience and molecular genetics, with earlier research focused on protein kinases, Alzheimer’s disease, neurodegeneration, and large scale biological data. I have also worked as an entrepreneur and in an early stage whole brain emulation company, where I contributed across research, strategy, governance, communications, partnerships, and operations.
My work in whole brain emulation led me to think more seriously about artificial sentience and digital minds. Technologies that model, copy, preserve, enhance, or instantiate aspects of human cognition raise questions not only about technical feasibility, but also about identity, consent, welfare, moral status, and governance.
My current interests include artificial sentience, neurotechnology, AI alignment, whole brain emulation, brain computer interfaces, cognitive enhancement, neuromorphic AI, and the transition to a post AGI society. I am particularly interested in proposals that aim to preserve human agency, intelligence, oversight capacity, or political relevance as AI systems become more capable.
As a mentor, I would be excited to help mentees turn broad and emerging questions into focused, decision relevant projects. I can support project scoping, literature synthesis, research design, writing, strategic analysis, and the development of public facing outputs.