Published March 30, 2026 | Version CC-BY-NC-ND 4.0
Journal article Open

LLMs are A Dead End in Search for General Machine Intelligence: A Review

Authors/Creators

  • 1. Department of Electrical Engineering, Shri Govindram Seksaria Institute of Technology and Science, Indore (Madhya Pradesh), India.

Contributors

Contact person:

  • 1. Department of Artificial Intelligence and Data Science, University of Mumbai, Mumbai (Maharashtra), India.
  • 2. Department of Electrical Engineering, Shri Govindram Seksaria Institute of Technology and Science, Indore (Madhya Pradesh), India.
  • 3. Department of Computer Science, University of Mumbai, Mumbai (Maharashtra), India.
  • 4. Department of Electrical Engineering, Shri Govindram Seksaria Institute of Technology and Science, Indore (Madhya Pradesh), India.
  • 5. Assistant Professor, Department of Electrical Engineering, Shri Govindram Seksaria Institute of Technology and Science, Indore (Madhya Pradesh), India.

Description

Abstract: This extensive review of large language models (LLMs) aims to highlight the importance of scaling the current generation of large language models toward artificial general intelligence, which is a dead end, while also considering the risks of unregulated use of such models. Through this, it is aimed to explicitly highlight the intelligence factor of current large language models and their malicious manipulative ability. While many large language model organisations compete to achieve better results by scaling up their models, this ultimately leads to the models' collapse. It is too early to understand the development and benefits of large language models; many have cited LLMs as the primary means of achieving general intelligence agents. To counter this, this paper gathers and evaluates resources from multiple research articles and tests several frequently used LLMs, highlighting their importance in different scenarios. As these models are trained on a wide variety of data, they exhibit domain-independent intelligent behaviour but fail to exhibit causal intelligent behaviour.

Files

F370715060126.pdf

Files (752.9 kB)

Name Size Download all
md5:01d5a4834efca618417d31a7dc95336d
752.9 kB Preview Download

Additional details

Identifiers

DOI
10.35940/ijsce.F3707.16010326
EISSN
2231-2307

Dates

Accepted
2026-03-15
Manuscript received on 18 December 2025 | First Revised Manuscript received on 08 January 2026 | Second Revised Manuscript received on 16 February 2026 | Manuscript Accepted on 15 March 2026 | Manuscript published on 30 March 2026

References

  • OpenAI et al., "GPT-4 Technical Report," arXiv preprint arXiv:2303.08774, Mar. 2023. [Online]. Available: https://arxiv.org/abs/2303.08774
  • Jones, C.R., Bergen, B.K.: Does GPT-4 pass the Turing test? (2024). https://arxiv.org/abs/2310.20216
  • S. Court and M. Elsner, "Shortcomings of LLMs for Low-Resource Translation: Retrieval and Understanding are Both the Problem," arXiv preprint arXiv:2406.15625, 2024. [Online]. Available: https://arxiv.org/abs/2406.15625
  • LeCun, Y.: A path towards autonomous machine intelligence version 0.9. 2, 2022-06-27. Open Review 62(1), 1–62 (2022). https://openreview.net/forum?id=BZ5a1r-kVsf
  • Wang, R., Todd, G., Xiao, Z., Yuan, X., Cˆot´e, M.-A., Clark, P., Jansen, P.: Can Language Models Serve as Text-Based World Simulators? (2024). https://arxiv.org/abs/2406.06485
  • Dubey, A., Jauhri, A., Pandey, A., Kadian, A., Al-Dahle, A., Letman, A., Mathur, A., Schelten, A., Yang, A., Fan, A., et al.: The llama 3 herd of models. arXiv preprint arXiv:2407.21783 (2024). https://arxiv.org/abs/2407.21783
  • Brown et al. (2020): Language Models are Few-Shot Learners. In: NeurIPS 2020 Proceedings. https://arxiv.org/abs/2005.14165
  • Kamoi, R., Zhang, Y., Zhang, N., Han, J., Zhang, R.: When Can LLMs Actually Correct Their Own Mistakes? A Critical Survey of Self Correction of LLMs (2024). https://arxiv.org/abs/2406.01297
  • Wei, J., Zhang, Y., Zhang, L.Y., Ding, M., Chen, C., Ong, K.-L., Zhang, J., Xiang, Y.: Memorisation in deep learning: A survey (2024). https://arxiv.org/abs/2406.03880
  • Blank, I.A.: What are large language models supposed to model? Trends in Cognitive Sciences 27(11), 987–989 (2023). DOI: https://doi.org/10.1016/j.tics.2023.08.006
  • Paech, S.J.: EQ-Bench: An Emotional Intelligence Benchmark for Large Language Models https://arxiv.org/abs/2312.06281
  • Nyamsuren, E., Taatgen, N.: Human reasoning module. Biologically Inspired Cognitive Architectures 8 (2014). DOI: https://doi.org/10.1016/j.bica.2014.02.002
  • Nasr, M., Carlini, N., Hayase, J., Jagielski, M., Cooper, A.F., Ippolito, D., Choquette-Choo, C.A., Wallace, E., Tram`er, F., Lee, K.: Scalable extraction of training data from (production) language models. ArXiv (2023). DOI: https://doi.org/10.48550/arXiv.2311.17035
  • Chollet, F.: On the measure of intelligence. http://arxiv.org/abs/1911.01547
  • Han, S.J., Ransom, K.J., Perfors, A., Kemp, C.: Inductive reasoning in humans and large language models. Cognitive Systems Research 83, 101155 (2024). DOI: https://doi.org/10.1016/j.cogsys.2023.101155
  • Houser, K.: LLMs are a dead end to AGI, says Franc¸ois Chollet (2024). https://www.freethink.com/robots-ai/arc-prize-agi
  • Opiel-ka, G., Rosenbusch, H., Vijverberg, V., Stevenson, C.E.: Do Large Language Models Solve ARC Visual Analogies Like People Do? (2024). https://arxiv.org/abs/2403.09734
  • Rinaldi, L., Karmiloff-Smith, A.: Intelligence as a developing function: A neuro-constructivist approach. Journal of Intelligence 5 (2017). https://www.mdpi.com/2079-3200/5/2/18
  • Fang, M., Deng, S., Zhang, Y., Shi, Z., Chen, L., Pechenizkiy, M., Wang, J.: Large language models are neurosymbolic reasoners. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 38, pp. 17985 17993 (2024). https://ojs.aaai.org/index.php/AAAI/article/view/29712
  • Wu, F., Zhang, N., Jha, S., McDaniel, P.D., Xiao, C.: A new era in LLM security: Exploring security concerns in real-world LLM-based systems. https://arxiv.org/abs/2402.18649
  • Yao, Y., Duan, J., Xu, K., Cai, Y., Sun, Z., Zhang, Y.: A survey on large language model (llm) security and privacy: The good, the bad, and the ugly. High-Confidence Computing, 100211 (2024) DOI: https://doi.org/10.1016/j.hcc.2024.100211
  • X. Chang, G. Dai, H. Di, and H. Ye, "Breaking the Prompt Wall (I): A Real-World Case Study of Attacking ChatGPT via Lightweight Prompt Injection." 2025. https://arxiv.org/abs/2504.16125
  • Narayanan, A.: Indirect prompt injection via hidden instructions on a webpage (2023). https://x.com/random_walker/status/1636923058370891778
  • Xu, Z., Liu, Y., Deng, G., Li, Y., Picek, S.: Llm jailbreak attack versus defence techniques–a comprehensive study. arXiv preprint arXiv:2402.13457 (2024). https://arxiv.org/abs/2402.13457
  • Shen, X., Chen, Z., Backes, M., Shen, Y., Zhang, Y.: "Do Anything Now": Characterising and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models (2024). https://arxiv.org/abs/2308.03825
  • Liu, X., Xu, N., Chen, M., Xiao, C.: AutoDAN: Generating stealthy jailbreak prompts on aligned large language models. In: The Twelfth International Conference on Learning Representations (2024). https://openreview.net/forum?id=7Jwpw4qKkb
  • Zou, A., Wang, Z., Carlini, N., Nasr, M., Kolter, J.Z., Fredrikson, M.: Universal and transferable adversarial attacks on aligned language models. arXiv preprint arXiv:2307.15043 (2023). https://arxiv.org/abs/2307.15043
  • Shumailov, I., Shumaylov, Z., Zhao, Y., Papernot, N., Anderson, R., Gal, Y.: AI models collapse when trained on recursively generated data: nature 631(8022), 755–759 (2024). DOI: https://doi.org/10.1038/s41586-024-07566-y
  • Huang, L., Yu, W., Ma, W., Zhong, W., Feng, Z., Wang, H., Chen, Q., Peng, W., Feng, X., Qin, B., et al.: A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. arXiv preprint arXiv:2311.05232 (2023). https://arxiv.org/abs/2311.05232
  • Cohen, S., Bitton, R., Nassi, B.: Here comes the AI worm: Unleashing zero-click worms that target genai-powered applications. ArXiv abs/2403.02817 (2024) https://arxiv.org/abs/2403.02817
  • Greshake, K., Abdelnabi, S., Mishra, S., Endres, C., Holz, T., Fritz, M.: Not what you've signed up for: Compromising Real-World LLM Integrated Applications with Indirect Prompt Injection (2023). https://arxiv.org/abs/2302.12173
  • Abdelaziz, I., Basu, K., Agarwal, M., Kumaravel, S., Stallone, M., Panda, R., Rizk, Y., Bhargav, G., Crouse, M., Gunasekara, C., Ikbal, S., Joshi, S., Karanam, H., Kumar, V., Munawar, A., Neelam, S., Raghu, D., Sharma, U., Soria, A.M., Sreedhar, D., Venkateswaran, P., Unuvar, M., Cox, D., Roukos, S., Lastras, L., Kapanipathi, P.: Granite-Function Calling Model: Introducing Function Calling Abilities via Multi-task Learning of Granular Tasks (2024). https://arxiv.org/abs/2407.00121
  • Chen, W., Li, Z., Ma, M.: Octopus: On-device language model for function calling of software APIs (2024). https://arxiv.org/abs/2404.01549
  • Wang, Y., Yu, J., Yao, Z., Zhang, J., Xie, Y., Tu, S., Fu, Y., Feng, Y., Zhang, J., Zhang, J., Huang, B., Li, Y., Yuan, H., Hou, L., Li, J., Tang, J.: A Solution-based LLM API-using Methodology for Academic Information Seeking (2024). https://arxiv.org/abs/2405.15165
  • Villalobos, P., Ho, A., Sevilla, J., Besiroglu, T., Heim, L., Hobbhahn, M.: Position: Will we run out of data? limits of LLM scaling based on human-generated data. In: Forty-first International Conference on Machine Learning (2024). https://openreview.net/forum?id=ViZcgDQjyG
  • Gerstgrasser, M., Schaeffer, R., Dey, A., Rafailov, R., Sleight, H., Hughes, J., Korbak, T., Agrawal, R., Pai, D., Gromov, A., Roberts, D.A., Yang, D., Donoho, D.L., Koyejo, S.: Is Model Collapse Inevitable? Breaking the Curse of Recursion by Accumulating Real and Synthetic Data (2024). https://arxiv.org/abs/2404.01413
  • Mart´ınez, G., Watson, L., Reviriego, P., Hern´andez, J.A., Juarez, M., Sarkar, R.: Towards Understanding the Interplay of Generative Artificial Intelligence and the Internet (2023). https://arxiv.org/abs/2306.06130
  • Zhang, Q., Zeng, B., Zhou, C., Go, G., Shi, H., Jiang, Y.: Human-imperceptible retrieval-poisoning attacks in LLM-powered applications. In: Companion Proceedings of the 32nd ACM International Conference on the Foundations of Software Engineering, pp. 502–506 (2024) DOI: https://dl.acm.org/doi/10.1145/3663529.3663793
  • Long, L., Wang, R., Xiao, R., Zhao, J., Ding, X., Chen, G., Wang, H.: On LLMs-driven synthetic data generation, curation, and evaluation: A survey. arXiv preprint arXiv:2406.15126 (2023) https://arxiv.org/abs/2406.15126
  • Yan, B., Li, K., Xu, M., Dong, Y., Zhang, Y., Ren, Z., Cheng, X.: On protecting the data privacy of large language models (llms): A survey. arXiv preprint arXiv:2403.05156 (2024). https://arxiv.org/abs/2403.05156
  • Inan, H., Upasani, K., Chi, J., Rungta, R., Iyer, K., Mao, Y., Tontchev, M., Hu, Q., Fuller, B., Testuggine, D., et al.: Llama guard: LLM-based input-output safeguard for human-ai conversations. arXiv preprint arXiv:2312.06674 (2023) https://arxiv.org/abs/2312.06674
  • Pal, M.: Meta faces backlash over WhatsApp jokes hurting religious sentiments. Times Now (2024) https://www.timesnownews.com/technology-science/meta-ai-faces backlash-on-whatsapp-jokes-hurting-religious-sentiments-article 109600087
  • Lukas, N., Salem, A., Sim, R., Tople, S., Wutschitz, L., Zanella B´eguelin, S.: Analysing leakage of personally identifiable information in language models. In: 2023 IEEE Symposium on Security and Privacy (SP), pp. 346–363 (2023). IEEE. DOI: https://doi.org/10.1109/SP46215.2023.10179418
  • He, F., Zhu, T., Ye, D., Liu, B., Zhou, W., Yu, P.S.: The emerging security and privacy of LLM agents: A survey with case studies. arXiv preprint arXiv:2407.19354 (2024). https://arxiv.org/abs/2407.19354
  • Majeed, A., Hwang, S.O.: Reliability issues of llms: ChatGPT, a case study. IEEE Reliability Magazine, 1–11 (2024). DOI: https://doi.org/10.1109/MRL.2024.3420849
  • Bender, E.M., Gebru, T., McMillan-Major, A., Shmitchell, S.: On the dangers of stochastic parrots: Can language models be too big? In: Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency. FAccT '21, pp. 610–623. Association for Computing Machinery, New York, NY, USA (2021). DOI: https://doi.org/10.1145/3442188.3445922
  • Arkoudas, K.: ChatGPT is no stochastic parrot. But it also claims that 1 is greater than 1. Philosophy & Technology 36(3), 54 (2023) DOI: https://doi.org/10.1007/s13347-023-00640-3
  • [48] Hicks, M.T., Humphries, J., Slater, J.: ChatGPT is bullshit. Ethics and Information Technology 26(2), 38 (2024). DOI: https://doi.org/10.1007/s10676-024-09702-3
  • Nejjar, M., Zacharias, L., Stiehle, F., Weber, I.: Llms for science: Usage for code generation and data analysis. arXiv preprint arXiv:2311.16733 (2023). https://arxiv.org/abs/2311.16733
  • He, Y., Wang, E., Rong, Y., Cheng, Z., Chen, H.: Security of AI agents. arXiv preprint arXiv:2406.08689 (2024). https://arxiv.org/abs/2406.08689
  • Hasani, R., Lechner, M., Wang, T.-H., Chahine, M., Amini, A., Rus, D.: Liquid structural state-space models. arXiv preprint arXiv:2209.12951 (2022) https://arxiv.org/abs/2209.12951