# Getting from Generative AI to Trustworthy AI: What LLMs might learn from Cyc

arXiv: https://arxiv.org/abs/2308.04445 · HTML: https://mlsanity.com/paper/2308.04445/

## Key facts

- Cited by 19 papers in the map (data through 14 Oct 2026).
- Builds on 2 of its 14 references that are in the map.
- First posted on arXiv 31 Jul 2023.

## Builds on (most cited first, 2 of 2)

- [Universal and Transferable Adversarial Attacks on Aligned Language Models](https://mlsanity.com/paper/2307.15043/) (2023)
- [How is ChatGPT's behavior changing over time?](https://mlsanity.com/paper/2307.09009/) (2023)

## Built on by (most cited first, 19 of 19)

- [Can Knowledge Graphs Reduce Hallucinations in LLMs? : A Survey](https://mlsanity.com/paper/2311.07914/) (2023)
- [On the Prospects of Incorporating Large Language Models (LLMs) in Automated Planning and Scheduling (APS)](https://mlsanity.com/paper/2401.02500/) (2024)
- [Neuro-Symbolic AI in 2024: A Systematic Review](https://mlsanity.com/paper/2501.05435/) (2025)
- [Towards Knowledge-driven Autonomous Driving](https://mlsanity.com/paper/2312.04316/) (2023)
- [Large Multi-Modal Models (LMMs) as Universal Foundation Models for AI-Native Wireless Systems](https://mlsanity.com/paper/2402.01748/) (2024)
- [Exploiting Language Models as a Source of Knowledge for Cognitive Agents](https://mlsanity.com/paper/2310.06846/) (2023)
- [Reasoning-as-Logic-Units: Scaling Test-Time Reasoning in Large Language Models Through Logic Unit Alignment](https://mlsanity.com/paper/2502.07803/) (2025)
- [Trustworthy AI: Deciding What to Decide](https://mlsanity.com/paper/2311.12604/) (2023)
- [Quo Vadis ChatGPT? From Large Language Models to Large Knowledge Models](https://mlsanity.com/paper/2405.19561/) (2024)
- [A Perspective on Large Language Models, Intelligent Machines, and Knowledge Acquisition](https://mlsanity.com/paper/2408.06598/) (2024)
- [Benchmarking graph construction by large language models for coherence-driven inference](https://mlsanity.com/paper/2502.13953/) (2025)
- [New contexts, old heuristics: How young people in India and the US trust online content in the age of generative AI](https://mlsanity.com/paper/2405.02522/) (2024)
- [Human-AI collectives produce the most accurate differential diagnoses](https://mlsanity.com/paper/2406.14981/) (2024)
- [VECSR: Virtually Embodied Common Sense Reasoning System](https://mlsanity.com/paper/2505.02144/) (2025)
- [Standardizing Knowledge Engineering Practices with a Reference Architecture](https://mlsanity.com/paper/2404.03624/) (2024)
- [What's in an embedding? Would a rose by any embedding smell as sweet?](https://mlsanity.com/paper/2406.06870/) (2024)
- [Building Trustworthy AI by Addressing its 16+2 Desiderata with Goal-Directed Commonsense Reasoning](https://mlsanity.com/paper/2506.12667/) (2025)
- [Coherence-driven inference for cybersecurity](https://mlsanity.com/paper/2509.18520/) (2025)
- [Adaptive Utility driven Resource Orchestration for Resilient AI (AURORA-AI)](https://mlsanity.com/paper/2606.27005/) (2026)

## Abstract (arXiv)

Generative AI, the most popular current approach to AI, consists of large language models (LLMs) that are trained to produce outputs that are plausible, but not necessarily correct. Although their abilities are often uncanny, they are lacking in aspects of reasoning, leading LLMs to be less than completely trustworthy. Furthermore, their results tend to be both unpredictable and uninterpretable. We lay out 16 desiderata for future AI, and discuss an alternative approach to AI which could theoretically address many of the limitations associated with current approaches: AI educated with curated pieces of explicit knowledge and rules of thumb, enabling an inference engine to automatically deduce the logical entailments of all that knowledge. Even long arguments produced this way can be both trustworthy and interpretable, since the full step-by-step line of reasoning is always available, and for each step the provenance of the knowledge used can be documented and audited. There is however a catch: if the logical language is expressive enough to fully represent the meaning of anything we can say in English, then the inference engine runs much too slowly. That's why symbolic AI systems typically settle for some fast but much less expressive logic, such as knowledge graphs. We describe how one AI system, Cyc, has developed ways to overcome that tradeoff and is able to reason in higher order logic in real time. We suggest that any trustworthy general AI will need to hybridize the approaches, the LLM approach and more formal approach, and lay out a path to realizing that dream.