How ethical was AI in 2021?

And more AI stories of the week

2021 made 2020 look like a prequel. But despite all its turbulence, it was still a memorable year for us at AIM -

• AIM was read by more than 23 million times in 2021

• Despite the pandemic, we worked with more than 100 brands on their marketing campaigns

• Added 40 new brands that worked with us

• Did 100+ virtual events that touched 50k+ data science & AI enthusiasts.

• Curated 2 custom conferences for leading tech brands

• Did half a dozen data science hiring hackathons

• Published more than 3500 articles & 100+ videos

• Our team doubled since the start of the year

• Created a new business unit called AIMrecruits, worked with more than 40+ brands on their data science hiring needs.

• Launched AIM Leaders Council, with 50+ members 

• MachineHack is now considered the leading competitive coding platform for ML developers

Signing out of 2021; excited for 2022. Wish you all a very happy new year.

---

1

How ethical was AI in 2021?

In a hopeful turn of events, 2021 witnessed big tech companies, governments and judiciaries finally take some of the most foundational steps to curb the growth of unethical AI.

We’ve curated a side-by-side of the year’s biggest ethical AI fiascos and the possible positives.

Also: How This AI Ethics Researcher Combines Anthropology And Technology To Build Human-First Solutions

+ DeepMind’s Scientist Sparks Fresh Debate Over AI Ethics

---

2

NVIDIA’s GauGAN2 vs OpenAI’s GLIDE

Just a few days back, openAI released GLIDE (Guided Language to Image Diffusion for Generation and Editing), a new 3.5 billion parameter text-to-image generation model that outperformed DALL·E.

Around a month back, NVIDIA, which has been actively working in this area for quite some time, revealed the sequel to its GauGAN model, the GauGAN2, which allows users to create real landscape photos.

Also: GauGAN2 From NVIDIA Includes A Text-To-Image Feature

+ OpenAI Brings Out GLIDE, Outperforms Its Own DALL-E

3

Meta’s Bean Machine

Bean Machine, Meta’s probabilistic programming system, is a PyTorch-based model to represent and learn about the uncertainties in ML models.

Still, in its beta phase, Bean Machine’s ‘uncertainty-aware’ learning algorithms allow the user to develop domain-specific probabilistic models and automatically identify its unobserved properties.

Also: Interesting Algorithms Released By Meta AI In 2021

+ Magic Pencil For Real: Meta AI’s Model To Bring Children’s Drawings To Life

---

4

Bluefield Vs Fungible Tussle

According to NVIDIA, its BlueField-2 DPU was able to achieve 41.5M IOPS during a storage benchmark test and claimed that it delivers “more than 4x IOPS than any other DPU.”

NVIDIA made this claim in response to Fungible’s announcement last month that Fungible’s 10M IOPS storage demonstration, which was conducted with the San Diego Supercomputing Center, set new records for storage initiator performance on NVMe and TCP.

Also: How Palo Alto Networks Is Boosting Cyber Defenses With NVIDIA’s BlueField DPU

+ NVIDIA Releases Eighth Generation Of Its Popular Conversational AI Software TensorRT

---

5

Featured Video: Back to office or WFH | Are we ready for the new normal.

---

6

Hands-on Guides for ML Developers

+ A Guide to Parallel and Distributed Deep Learning for Beginners

+ LUDWIG: A Type-Based Declarative Deep Learning Toolbox

+ A Beginner’s Guide to Text Classification using BERT Features

+ How Fast Fourier Convolution Can Replace the Convolutional Layer of CNN?

+ What is Extreme Multilabel Text Classification?

---

7

People & Startups

+ XAI Is Critical To Ensure Transparency: Sreekanth Menon, Genpact

+ Identifying And Retaining Good Talent Is Crucial In Data Analytics: Vineet Shukla, Mahindra Group

+ Edelweiss’ Head of Investment Management Anshu Kapoor On Indian Tech Companies Foraying IPO Route

+ Lightelligence's PACE: a major breakthrough in optical computing

+ How this Bengaluru based company did sentiment analysis for Harry Potter Reunion teaser

---

8

“Chinese AI Dragon” goes public despite blacklisting

Chinese artificial intelligence startup SenseTime’s initial Hong Kong public offering plans to offer 1.5 billion shares to raise a total of US $741 million. The trading of shares is set to start on December 30.

This comes despite being blacklisted by the US government a few weeks back, which was supposed to derail this IPO process. It is now the largest AI software firm in Asia, with an estimated 11% market share and a market value of around US $16 billion.

Also: Misuse of Artificial Intelligence in China

+ Questioning The Future Of Africa’s Digital Sovereignty, Powered By China

---

9

Interesting Innovations From DeepMind & OpenAI In 2021

Innovation leader DeepMind put its best foot forward this year and released some mind-blowing systems with the potential to create a big impact.

Not even a decade old, OpenAI has made a name for itself as a leading AI research lab worldwide. It gave the world GPT-3 in 2020 – a path-breaking innovation that uses deep learning to give us a human-like text. This year, too, saw OpenAI testing its limits and continuing its streak in bringing out algorithms and models that can create a massive impact.

Also: A historical tale of DeepMind’s games

+ Kiss Me, Transformer – A Journey of OpenAI

---

10

BOTTOM OF THE NEWS

Here's what all happened last week.

+ Microsoft open-sources distributed ML library SynapseML

+ Sashastra Seema Bal wants AI-based tech solution to guard India's borders

+ Indian Army establishes new Quantum Computing Lab and AI centre

+ IIT-Delhi introduces 'Pillay Chair Professor' to promote research in ML

+ Bengaluru has the highest AI talent pool outside the US, ranked 5th in a study