Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Aiming beyond slight increases in accuracy

Nature Reviews Chemistry volume 7, pages 227–228 (2023)Cite this article

Subjects

Owing to the diminishing returns of deep learning and the focus on model accuracy, machine learning for chemistry might become an endeavour exclusive to well-funded institutions and industry. Extending the focus to model efficiency and interpretability will make machine learning for chemistry more inclusive and drive methodological progress.

Deep learning is a machine learning technique based on layered artificial neural networks that was pioneered by Alexey Grigorevich Ivakhnenko in the 1960s. Catalysed by computer hardware advances during the past two decades, deep learning has more recently found broad application in virtually every scientific field, including chemistry1. Different variants of deep learning are being used to generate new molecules, predict the retrosynthetic pathways of biocatalysed reactions, and improve predictions of physical and chemical properties of materials based on density functional theory calculations2,3.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Rent or buy this article

Get just this article for as long as you need it

$39.95

Learn more

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: More than accuracy.

References

  1. Ivakhnenko, A. G. The group method of data handling, a rival of the method of stochastic approximation. Sov. Automat. Contr. 13, 43–55 (1968).

    Google Scholar 

  2. Keith, J. A. et al. Combining machine learning and computational chemistry for predictive insights into chemical systems. Chem. Rev. 121, 9816–9872 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Probst, D. et al. Biocatalysed synthesis planning using data-driven learning. Nat. Commun. 13, 964 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Thompson, N. C., Greenewald, K., Lee, K. & Manso, G. F. Deep learning’s diminishing returns. IEEE Spectr. 58, 50–55 (2021).

    Article  Google Scholar 

  5. Ahmed, N. & Wahed, M. The de-democratization of AI: deep Learning and the compute divide in artificial intelligence research. Preprint at https://arxiv.org/abs/2010.15581 (2020).

  6. Jurowetzki, R., Hain, D., Mateos-Garcia, J. & Stathoulopoulos, K. The privatization of AI research(-ers): causes and potential consequences–from university-industry interaction to public research brain-drain? Preprint at https://arxiv.org/abs/2102.01648 (2021).

  7. Schwartz, R., Dodge, J., Smith, N. A. & Etzioni, O. Green AI. Commun. ACM 63, 54–63 (2020).

    Article  Google Scholar 

  8. Patterson, D. et al. The carbon footprint of machine learning training will plateau, then shrink. Preprint at https://arxiv.org/abs/2204.05149 (2022).

  9. Probst, D. Social and environmental impact of recent developments in machine learning on biology and chemistry research. Preprint at https://arxiv.org/abs/2210.00356 (2022).

  10. Scao, T. L. et al. BLOOM: a 176B-parameter open-access multilingual language model. Preprint at https://arxiv.org/abs/2211.05100 (2022).

Download references

Acknowledgements

The author thanks Pierre Vandergheynst for hosting them as a postdoctoral researcher at the École Polytechnique Fédérale de Lausanne.

Author information

Authors and Affiliations

  1. School of Engineering, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Vaud, Switzerland

    Daniel Probst

Authors
  1. Daniel Probst
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Daniel Probst.

Ethics declarations

Competing interests

The author declares no competing interests.

Rights and permissions

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Probst, D. Aiming beyond slight increases in accuracy. Nat Rev Chem 7, 227–228 (2023). https://doi.org/10.1038/s41570-023-00480-3

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41570-023-00480-3

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing