Since Deeplearning born, AI has taken a great progress in developing, and compare with hope of mankinds, it seems that AI developing speed is more slowly. AI winter is well on its way, is it true as Filip Piekiewski conclusion？
Even if it turned out to be true, it will not be the first winter for AI. The last one was in the late 1990s, a lasting ice age for multilayer ANNs that were deemed unstable in terms of learning convergence... These techniques have nevertheless made their way ever since. And nowadays, really with very strong multidisciplinary convergences, AI techniques (among which ANNs) will undoubtedly experience new future beautiful spring seasons, as advanced tools in the service of human beings... Every spring marks an increase in maturity and knowledge in these fields.
To your question why, answers by referring non-exhaustively to:
- A. Irpan, " Deep Reinforcement Learning Doesn't Work Yet ", 2018 -https://www.alexirpan.com/2018/02/14/rl-hard.html
- Akhtar et al., " Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Survey ", 2018 -
Article Threat of Adversarial Attacks on Deep Learning in Computer V...
- Regneri et al., "Analyzing Hypersensitive AI: Instability in Corporate-Scale Machine Learning ", 2018 -
Preprint Analyzing Hypersensitive AI: Instability in Corporate-Scale ...
- Antun et al., "On instabilities of deep learning in image reconstruction - Does AI come at a cost? ", 2019 -
Preprint On instabilities of deep learning in image reconstruction - ...
there is always something new coming up, AI will never have a long winter!
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence