How Fast Do Algorithms Improve? [Point of View]

Sherry, Yash; Thompson, Neil

doi:10.1109/jproc.2021.3107219

Cited by 13 publications

(5 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…3D printing), AI can greatly enlarge customization possibilities and provide consumers with unique items tailored to their needs. The ability to nurture deep learning processes with data and obtain constant algorithmic improvements throughout time will not only streamline existing production processes and facilitate the search for quicker solutions to extant problems (Sherry & Thompson, 2021) but it will also augment the potential for consumers' profilization (see the recent campaign by Ferrero with ‘Unique Nutella’). In other words, AI increases the potential for the absorptive capacity of both solution knowledge and need knowledge (Schweisfurth & Raasch, 2018).…”

Section: The Multilevel Implications Of Ai1mentioning

confidence: 99%

Creativity and artificial intelligence: A multilevel perspective

Grilli,

Pedota

2024

Creat Innov Manage

View full text Add to dashboard Cite

Artificial intelligence is likely to revolutionize multiple aspects of organizational creativity. Through a multilevel theoretical lens, the present paper reviews the extant body of knowledge on creativity at individual, team and organizational levels, and draws a series of propositions on how the implementation of artificial intelligence may affect each level. Spanning cognitive, behavioural and psychological domains, our propositions aim at directing future research efforts on important creativity‐related areas likely to be affected by artificial intelligence, including the trade‐off between convergent and divergent thinking, the distribution of skills within groups, and the absorptive capacity of organizations.

show abstract

Section: The Multilevel Implications Of Ai1mentioning

confidence: 99%

Creativity and artificial intelligence: A multilevel perspective

Grilli,

Pedota

2024

Creat Innov Manage

View full text Add to dashboard Cite

show abstract

“…Moreover, progress in computing performance over the past decades has been driven to a non-negligible part by improvements in algorithm design, often inspired by a better understanding of some of the key principles behind the workings of the brain. The exponential development and use of neural networks, for instance, was responsible for vast improvements over and above what simple hardware developments would have made possible (Sherry and Thompson, 2021 ).…”

Section: Understanding the Mechanisms Of The Trilemmamentioning

confidence: 99%

The AI trilemma: Saving the planet without ruining our jobs

Ernst

2022

Front. Artif. Intell.

View full text Add to dashboard Cite

Digitalization and artificial intelligence increasingly affect the world of work. Rising risk of massive job losses have sparked technological fears. Limited income and productivity gains concentrated among a few tech companies are fueling inequalities. In addition, the increasing ecological footprint of digital technologies has become the focus of much discussion. This creates a trilemma of rising inequality, low productivity growth and high ecological costs brought by technological progress. How can this trilemma be resolved? Which digital applications should be promoted specifically? And what should policymakers do to address this trilemma? This contribution shows that policymakers should create suitable conditions to fully exploit the potential in the area of network applications (transport, information exchange, supply, provisioning) in order to reap maximum societal benefits that can be widely shared. This requires shifting incentives away from current uses toward those that can, at least partially, address the trilemma. The contribution analyses the scope and limits of current policy instruments in this regard and discusses alternative approaches that are more aligned with the properties of the emerging technological paradigm underlying the digital economy. In particular, it discusses the possibility of institutional innovations required to address the socio-economic challenges resulting from the technological innovations brought about by artificial intelligence.

show abstract

“…Theory tells us that the lower bound for the computational intensity of regularized flexible models is O(P erf ormance 4 ), which is much better than current deep learning scaling. Encouragingly, there is historical precedent for algorithms improving rapidly [79].…”

Section: Futurementioning

confidence: 99%

The Computational Limits of Deep Learning

Thompson¹,

Greenewald²,

Lee³

et al. 2020

Preprint

Self Cite

115

View full text Add to dashboard Cite

Deep learning's recent history has been one of achievement: from triumphing over humans in the game of Go to world-leading performance in image recognition, voice recognition, translation, and other tasks. But this progress has come with a voracious appetite for computing power. This article reports on the computational demands of Deep Learning applications in five prominent application areas and shows that progress in all five is strongly reliant on increases in computing power. Extrapolating forward this reliance reveals that progress along current lines is rapidly becoming economically, technically, and environmentally unsustainable. Thus, continued progress in these applications will require dramatically more computationally-efficient methods, which will either have to come from changes to deep learning or from moving to other machine learning methods.

show abstract

How Fast Do Algorithms Improve? [Point of View]

Abstract: This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

Cited by 13 publications

References 12 publications

Creativity and artificial intelligence: A multilevel perspective

Creativity and artificial intelligence: A multilevel perspective

The AI trilemma: Saving the planet without ruining our jobs

The Computational Limits of Deep Learning

Contact Info

Product

Resources

About