SiC26 | Superintelligence Conference

Motivation

Irving J., Good defines Superintelligence in 1965 as: “Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever.” The recent developments in Neural Networks with Transformers and Large Language Models are showing us glimpses of what Superintelligence is capable of, affecting every field of knowledge and research. It is essential to prepare as well as understand the foundations of such machines. The topic of Superintelligence is interdisciplinary in nature.

Questions

• What is Superintelligence? Metrics, tests, levels and evaluations
• How to solve AGI?
• Theoretical foundations of Superintelligence
• Ethical, social and societal questions of AI
• AI Alignment
• Vibe coding
• Consciousness and AI
• Brain-computer interfaces
• Cognitive science and AI reasoning
• Neural Network architectures, Transformers and Large Language Models.

Call for Papers

The Organising Committee of the Superintelligence Conference (SiC26) invites researchers, academics, and industry professionals to submit original papers on all aspects of superintelligence. SiC26 aims to bring together leading interdisciplinary experts to discuss the latest advancements, challenges, and ethical considerations in the development and implications of superintelligent systems. SIC26 will be held at the University of Exeter (UK), 9-11th September 2026.

There is no agreed-upon definition of superintelligence (SI), we may think of SI as an artificial entity that surpasses humans in overall intelligence or in some particular measure of intelligence. SiC intends to become a forum to discuss matters of superintelligence and define it. Recent developments in neural networks with transformers and large language models (LLM) are showing us glimpses of what superintelligence is capable of, affecting every field of knowledge and research. It is essential to prepare as well as understand the foundations of such machines.

The topic of SI is interdisciplinary in nature. We welcome submissions from a range of fields including: computer science and engineering (e.g. robotics, expert systems, natural language processing, artificial intelligence, human-computer interaction), philosophy and ethics, social sciences (e.g. political science, psychology, economics), environmental and space sciences, law and public policy, humanities(literature, art and history), neuroscience and biomedical sciences (e.g. animal behaviour, cognition).

Topics of interest (conference streams) include but are not limited to:

• Pathways and foundations of SI (e.g. scaling laws and their limits, post-Transformer architectures [state-space models, JEPA and world models, neuro-symbolic systems], training beyond pretraining [RLHF/RLAIF/RLVR, self-play, synthetic data], agency and long-horizon capability, self-improvement and automated ML research, embodiment and sensorimotor grounding, biological and hybrid pathways [brain emulation, organoid intelligence, BCIs], theoretical and information-theoretic limits of general intelligence)
• Conceptualising and evaluating SI (e.g. defining SI, AI’s progress towards SI [from LLMs to true understanding], uncertainty in AI and SI, quantum computing and its role in SI, quantifying SI [metrics, tests, levels, evaluations], impacts of commercial LLMs on the future of AI, explainability and interpretability in SI systems)
• Cognitive science of SI (autonomous systems and decision-making at SI levels, AI reasoning, biological vs. artificial pathways to SI, neuroscience-inspired approaches to SI, human-AI collaboration and augmented intelligence)
• Ethical, legal, and policy implications of SI (e.g. existential risk mitigation, the role of embodiment in SI agents, adversarial AI and security threats, consciousness of SI agents, machine morality, frameworks for SI governance)
• Frontiers and human progress with SI (e.g. AI Scientist, SI in science discovery and acceleration, computational creativity and SI art generation, how SI changes human life routines and re-shapes civilisations, multi-agent coordination among SI entities, role of researchers in a world with SI, democracy and sustainable development)

Submission guidelines

Authors are invited to submit original academic works in the following formats:

• Workshop proposal (1-page)
• Abstract submission (1-page including references)
• Research poster (1 slide)
• Full paper (up to 4 pages, including references)

All submissions must use the “MDPI article template”: https://www.overleaf.com/latex/templates/mdpi-article-template/fcpwsspfzsph. Submissions must be original (unpublished) work. Submissions will be peer-reviewed by an expert committee and accepted full papers will be published in the conference proceedings. Reviews will be double-blind, hence do not include your names as authors for your submission, you may use “anonymous”.

Authors should ensure to create their profile on Openreview 2 weeks ahead of the submission deadline.

Important dates

• Submission Deadlines: 5th July 2026 (all submission formats)
• Notification of Acceptance: 15th August 2026
• Camera Ready : 1st September 2026
• Conference Dates: 9-11th September 2026

Source and more info: https://www.superintelligenceconference.org/call-for-papers