And yes, Lush/SN used a homegrown Lisp interpreter, to which a compiler was added in the early 1990s.
@ylecun
-
Better terminology for multidimensional arrays in machine learning
By
–
Do you have a better name than "multidimensional array" though? The name tensor is convenient, if mathematically inaccurate.
-
Tensor Engine History: From Bell Labs to PyTorch
By
–
The tensor engine was first implemented inside SN3 (before it was called Lush) in 1992 at Bell Labs by Léon Bottom and me.
The naming convention has survived to this day in PyTorch and other libraries. The naming of the tensor operations was reused in EBlearn (C++ deep learning -
Balancing Tech Optimism with Democracy and Risk Management
By
–
On peut être à la fois techno-optimiste et anti-techno-féodaliste, pro-démocratie, conscient des risques et travaillant à maximiser les effets bénéfiques et à minimiser les effets négatifs.
-
European vs American government trust and AI policy implications
By
–
I don't pretend to speak for everyone, but there is no question that, on average, Europeans trust their government (and expect more from it) than Americans.
-
Nexar BADAS 2.0: V-JEPA collision prediction system
By
–
BADAS 2.0: new collision prediction system from Nexar.
It's based on V-JEPA 2!!
JEPA is going to save lives. -
EU Privacy Laws Contrast Sharply with American Regulatory Approach
By
–
EU politics is very different from US politics.
Unlike the US, the EU actually has strong privacy protection laws, including for private data (protections that American companies have often complained about), including for protections from intrusion from government as well as -
Library Destruction and Knowledge Preservation in Digital Age
By
–
Destroying a library brings the dark ages. https://t.co/RjBjKVWjms
— Yann LeCun (@ylecun) 16 avril 2026Destroying a library brings the dark ages.
-
Republican Anti-Science Policies Drive Academic Brain Drain
By
–
Maybe if the Republican party hadn't consistently been anti-science, anti-intellectual, anti-academic freedom, hadn't imposed ridiculous restrictions on visas for foreign scientists and graduates, and hadn't promoted obscurantist policies academics would not have abandoned it.
-
Rethinking Causal Inference and Technology Benefit Assessment Methods
By
–
You should reexamine your causal inference procedure. Also your fact collection methods. Also your evaluation of the benefits of technology.