Today's Tech Digest - Feb 02, 2020

Just how big a deal is Google’s new Meena chatbot model?

Meena can chat, over a few turns of a conversation, believably. Meena, however, cannot reliably teach you anything. Meena is not trying to help you finish a task or learn something new specifically. It converses with no explicit goal or purpose. While we probably spend too much of our time chatting about not much of importance, we tend to be looking for something specific when interacting with a bot-powered digital service. We want to get a ticket booked or a customer support issue resolved. Or we want to get accurate information about a particular domain or emotional or psychological support for a challenge we are facing. Conversational products have a purpose, and even if they fail at the more open-ended questions, they are trying to work with you to complete a task. Meena places the human-likeness of the conversation above all. However, there is much for us to learn about what is an appropriate conversational approach given different types of tasks. There is research that shows that more “robot” like responses are preferable in certain situations (especially where sensitive personal information is involved) and that being human-like is not the end-all and be-all of bots. Where does Meena, with the conversations it has learned from social media interactions, find a role?

How bacteria could run the Internet of Things

Environmental IoT is one area they say could benefit. In smart cities, for example, bacteria could be programmed to sense for pollutants. Microbes have good chemical-sensing functions and could turn out to work better than electronic sensors. In fact, the authors say that microbes share some of the same sensing, actuating, communicating and processing abilities that the computerized IoT has. In the case of sensing and actuating, bacteria can detect chemicals, electromagnetic fields, light, mechanical stress and temperature — just what’s required in a traditional printed circuit board-based sensor. Plus, the microbes respond. They can produce colored proteins, for example. And not only that, they respond in a more nuanced way compared to the chip-based sensors. They can be more sensitive, as one example. ... Bacteria should become a “substrate to build a biological version of the Internet of Things,” the scientists say. Interestingly, similar to how traditional IoT has been propelled forward by tech hobbyists mucking around with Arduino microcontrollers and Raspberry Pi educational mini-computers, Kim and Posland reckon it will be do-it-yourself biology that will kick-start IoBNT.

AI still doesn’t have the common sense to understand human language

The test was originally designed with the idea that such problems couldn’t be answered without a deeper grasp of semantics. State-of-the-art deep-learning models can now reach around 90% accuracy, so it would seem that NLP has gotten closer to its goal. But in their paper, which will receive the Outstanding Paper Award at next month’s AAAI conference, the researchers challenge the effectiveness of the benchmark and, thus, the level of progress that the field has actually made. They created a significantly larger data set, dubbed WinoGrande, with 44,000 of the same types of problems. To do so, they designed a crowdsourcing scheme to quickly create and validate new sentence pairs. (Part of the reason the Winograd data set is so small is that it was hand-crafted by experts.) Workers on Amazon Mechanical Turk created new sentences with required words selected through a randomization procedure. Each sentence pair was then given to three additional workers and kept only if it met three criteria: at least two workers selected the correct answers, all three deemed the options unambiguous, and the pronoun’s references couldn’t be deduced through simple word associations.

A new bill could punish web platforms for using end-to-end encryption

The bill doesn’t lay out specific rules. But the committee — which would be chaired by the Attorney General — is likely to limit how companies encrypt users’ data. Large web companies have moved toward end-to-end encryption (which keeps data encrypted for anyone outside a conversation, including the companies themselves) in recent years. Facebook has added end-to-end encryption to apps like Messenger and Whatsapp, for example, and it’s reportedly pushing it for other services as well. US Attorney General William Barr has condemned the move, saying it would prevent law enforcement from finding criminals, but Facebook isn’t required to comply. Under the EARN IT Act, though, a committee could require Facebook and other companies to add a backdoor for law enforcement. Riana Pfefferkorn, a member of the Stanford Law School’s Center for Internet and Society, wrote a detailed critique of the draft. She points out that the committee would have little oversight, and the Attorney General could also unilaterally modify the rules. The Justice Department has pushed encryption backdoors for years, citing threats like terrorism, but they haven’t gotten legal traction. Now, encryption opponents are riding the coattails of the backlash against big tech platforms and fears about child exploitation online.

Technologies of the future, but where are AI and ML headed to?

The fluid nature of data science allows people from multiple fields of expertise to come and crack it. Shantanu believes if JRR Tolkien, being the brilliant linguist that he was, pursued data science to develop NLP models, he would have been the greatest NLP expert ever, and that is the kind of liberty and scope data science offers. ... For a country like India, acquiring new skills is not something of a luxury but a necessary requirement, and the trends of upskilling and reskilling are also currently on the rise to complement with the same. But data science, machine learning, and artificial intelligence are those fields where mere book-reading and formulaic interpretation and execution just does not cut it. If one aspires to have a competitive career in futuristic technologies, machine learning and data science have a larger spectrum of required understanding of probability, statistics, and mathematics on a fundamental level. To break the myths around programmers and software developers entering this market, machine learning involves understanding of basic programming languages (Python, SQL, R), linear algebra and calculus, as well as inferential and descriptive statistics.

Read more here ...