One Hundred Year Study on Artificial Intelligence, or AI100

A newly published report on the state of artificial intelligence says the field has reached a turning point where attention must be paid to the everyday applications and even abuses of AI technology

“In the past five years, AI has made the leap from something that mostly happens in research labs or other highly controlled settings to something that’s out in society affecting people’s lives,” Brown University computer scientist Michael Littman, who chaired the report panel, said in a news release.

“That’s really exciting, because this technology is doing some amazing things that we could only dream about five or ten years ago,” Littman added. “But at the same time the field is coming to grips with the societal impact of this technology, and I think the next frontier is thinking about ways we can get the benefits from AI while minimizing the risks.”

Those risks include deep-fake images and videos that are used to spread misinformation or harm people’s reputations; online bots that are used to manipulate public opinionalgorithmic bias that infects AI with all-too-human prejudices; and pattern recognition systems that can invade personal privacy by piecing together data from multiple sources.

The report says computer scientists must work more closely with experts in the social sciences, the legal system and law enforcement to reduce those risks.

References:

What is NLP (Natural Language Processing)?

0:00 – Intro
0:38 – Unstructured data
1:12 – Structured data
2:03 – Natural Language Understanding (NLU) & Natural Language Generation (NLG)
2:36 – Machine Translation use case
3:40 – Virtual Assistance / Chat Bots use case
4:14 – Sentiment Analysis use case
4:44 – Spam Detection use case
5:44 – Tokenization
6:18 – Stemming & Lemmatization
7:42 – Part of Speech Tagging
8:22 – Named Entity Recognition (NER)
9:08 – Summary

Intel Add AI/ML Improvements to Sapphire Rapids with AMX

The full article is taken from With AMX, Intel Adds AI/ML Sparkle to Sapphire Rapids

Picture taken from With AMX, Intel Adds AI/ML Sparkle to Sapphire Rapids

The best way for now to think of AMX is that it’s a matrix math overlay for the AVX-512 vector math units, as shown below. We can think of it like a “TensorCore” type unit for the CPU. The details about what this is were only a short snippet of the overall event, but it at least gives us an idea of how much space Intel is granting to training and inference specifically.

Data comes directly into the tiles while at the same time, the host hops ahead and dispatches the loads for the toles. TMUL operates on data the moment it’s ready. At the end of each multiplication round, the tiles move to cache and SIMD post-processing and storing. The goal on the software side is to make sure both the host and AMX unit are running simultaneously.

The prioritization for AMX toward real-world AI workloads also meant a reckoning for how users were considering training versus inference. While the latency and programmability benefits of having training stay local are critical, and could well be a selling point for scalable training workloads on the CPU, inference has been the sweet spot for Intel thus far and AMX caters to that realization.

From The Next Platform “With AMX, Intel Adds AI/ML Sparkle to Sapphire Rapids”

NVIDIA Special Address at SIGGRAPH 2021

NVIDIA and SIGGRAPH share a long history of innovation and discovery. Over the last 25 years our community has seen giant leaps forward, driven by brilliant minds and curious explorers. We are now upon the opening moments of an AI-powered revolution in computer graphics with massive advancements in rendering, AI, simulation, and compute technologies across every industry. With open standards and connected ecosystems, we are on the cusp of achieving a new way to interact and exist with graphics in shared virtual worlds.

Quantum Computing can revolutionize AI

Quantum computers can process complex information at a mind-boggling speed and should eventually vastly outperform even the most powerful of today’s conventional computers. This includes the rapid training of machine learning models and the creation of optimized algorithms. Years of analysis can be cut to a short time with an optimized and stable AI that is powered by quantum computing. The combined solution is expected to bring changes to the AI hardware ecosystem

Techhq.com “Why AI will be so core to real-world quantum computing”

In a report by McKinsey, quantum computers have four fundamental capabilities that differentiate them from today’s classical computers: quantum simulation, in which quantum computers model complex molecules; optimization (that is, solving multivariable problems with unprecedented speed); quantum artificial intelligence (AI), utilizes better algorithms that could transform machine learning across industries as diverse as pharma and automotive; and prime factorization, which could revolutionize encryption.

Techhq.com “Why AI will be so core to real-world quantum computing”

For more information, do take a look at Why AI will be so core to real-world quantum computing

Images Taken from https://sociable.co/technology/could-quantum-computing-and-exotic-materials-facilitate-ai-human-cyborgs/

Performance Required for Deep Learning

There is this question that I wanted to find out about deep learning. What are essential System, Network, Protocol that will speed up the Training and/or Inferencing. There may not be necessary to employ the same level of requirements from Training to Inferencing and Vice Versa. I have received this information during a Nvidia Presentation

Training:

  1. Scalability requires ultra-fast networking
  2. Same hardware needs as HPC
  3. Extreme network bandwidth
  4. RDMA
  5. SHARP (Mellanox Scalable Hierarchical Aggregation and Reduction Protocol)
  6. GPUDirect (https://developer.nvidia.com/gpudirect)
  7. Fast Access Storage

Influencing

  1. Highly Transactional
  2. Ultra-low Latency
  3. Instant Network Response
  4. RDMA
  5. PeerDirect, GPUDirect