Summary: Our starting assumption that sequence problems (language, speech, and others) are the natural domain of RNNs is being challenged. Temporal Convolutional Nets (TCNs) which are our workhorse CNNs with a few new features are outperforming RNNs on major applications today. Looks like RNNs may well be history. It's only been since 2014 or 2015 when our DNN-powered applications passed the 95% accuracy point on text and speech recognition allowing for whole generations of chatbots, personal assistants, and instant translators. Convolutional Neural Nets (CNNs) are the acknowledged workhorse of image and video recognition while Recurrent Neural Nets (RNNs) became the same for all things language. One of the key differences is that CNNs can recognize features in static images (or video when considered one frame at a time) while RNNs exceled at text and speech which were recognized as sequence or time-dependent problems.

Prediction #6: The U.S. federal government will adopt a more proactive policy approach to AI in 2021 ... [ ] under President Biden. Below are 10 bold predictions about what will unfold in the world of artificial intelligence in 2021, from academic research to startups to capital markets to regulation. To keep ourselves honest, we will revisit these predictions in December 2021 to grade how we did. Autonomous vehicle developers like Waymo and Cruise have massive ongoing cash needs. Public market investors are thirsty for IPOs.

MOLOCO, an adtech startup that uses machine learning to build mobile campaigns, announced today it has raised $150 million in new Series C funding led by Tiger Global Management, taking its valuation to $1.5 billion. This is separate from the $20 million Series C round MOLOCO announced three months ago, which brought it to unicorn status. Co-founder and chief executive officer Ikkjin Ahn told TechCrunch that MOLOCO raised again so soon because "as we gear up for a potential IPO, we wanted more funding to help us grow faster." Founded in 2013 and based in Redwood City, California, MOLOCO has now raised $200 million in total. The company claims it has "consistently grown in excess of 100% annually," and has an annual net revenue run rate of more than $100 million.

This week, we learned a lot more about the inner workings of AI fairness and ethics operations at Facebook and Google and how things have gone wrong. On Monday, a Google employee group wrote a letter asking Congress and state lawmakers to pass legislation to protect AI ethics whistleblowers. That letter cites VentureBeat reporting about the potential policy outcomes of Google firing former Ethical AI team co-lead Timnit Gebru. It also cites research by UC Berkeley law professor Sonia Katyal, who told VentureBeat, "What we should be concerned about is a world where all of the most talented researchers like [Gebru] get hired at these places and then effectively muzzled from speaking. And when that happens, whistleblower protections become essential."

As Valentine's Day approaches, couples across the country are preparing for this long-standing tradition--and there's a very good chance they met through online dating. But while dating apps can help people find a partner (or just a fun date), they can also subject users to incredible hate and harassment. Despite the fact that dating apps have accrued significant reach and influence, these companies provide very little transparency around how they keep users safe and how they moderate content. Much of the conversation around online platform accountability focuses on companies like Facebook and Google. But dating apps face many of the same issues.

Prediction #6: The U.S. federal government will adopt a more proactive policy approach to AI in 2021 ... [ ] under President Biden. Below are 10 bold predictions about what will unfold in the world of artificial intelligence in 2021, from academic research to startups to capital markets to regulation. To keep ourselves honest, we will revisit these predictions in December 2021 to grade how we did. Autonomous vehicle developers like Waymo and Cruise have massive ongoing cash needs. Public market investors are thirsty for IPOs.

The present world is overwhelmed with digital technologies: computers, cell phones, as well as our cooker, washing machine, vehicle, and so on. In 1998 the principal standard computer was dispatched into the market: the Macintosh 128k. Today, 17 years later, existence without such computers is incomprehensible. We even ceaselessly carry one in our pocket: the cell phone. In close to no time, digital technology has prevailed in thoroughly adapting our lifestyle and flipping around it.

Many of us have had the feeling that technology, which continues to change at an ever-dizzying pace, may be leaving us behind. That was embodied this past week during a Congressional hearing, nominally convened to investigate antitrust concerns of four big tech titans: Amazon, Apple, Facebook and Google. While the five-and-a-half-hour inquiry touched on a range topics from pesky spam filters and search results to how companies approached acquisitions, the House Judiciary subcommittee hearing laid one thing bare: A sizable disconnect appears to exist between the technology Americans are using and depending on in their daily lives and the knowledge base of people with the power and responsibility to decide its future and regulation. "Consumers and investors walk away feeling like a lot of these lawmakers don't really understand the business models to an extent that they could then navigate them and put laws in place that will dictate the future of where they go," said Daniel Ives, an analyst with Wedbush Securities. The antitrust subcommittee hearing had been convened to look into the tech giants' market dominance.

The UK's Competition and Markets Authority (CMA) has called on the UK government to create "a new pro-competition regulatory regime" that can control Facebook, Google and other technology companies that are primarily funded by digital advertising. The non-ministerial department has completed a study announced last July and concluded that "existing laws are not suitable for effective regulation." To combat the problem, it's recommending that a new Digital Markets Unit be set up with major oversight and powers. The Unit was first proposed in a report published by the Digital Competition Expert Panel (DCEP) -- a group chaired by Professor Jason Furman, a former chief economist when Barack Obama was president -- in March 2019. The CMA believes it should have a code of conduct that ensures Facebook and Google don't veer into "exploitative or exclusionary practices," or do anything that is likely to reduce public trust and transparency.

Great to have you on the show, Ali. Thank you, Stephan, for having me here. It's a real pleasure to have me here. I listen to your podcast. It's great to be on it. It's great to have you on it. I've been trying to get you on this show for a while now. I'm really happy to have that finally come to pass. I'd love for our listeners to understand a bit more about the under the hood of Trendy Butler. How is this powered by an AI where all your competitors–the different subscription box companies–they seem to rely on human stylists, and you've got an AI? Most of them use a lot of AI to make their decision making. Except with us, we rely heavily on our AI. Obviously, there is still a stylist because as much as you could throw as many AIs as you want on to this, still there's an element of creativity that's needed for a person to be applying to this. You never could control, let's say the system picking an orange short with a purple shirt, you got to sometimes stop it. Most of it is actually done through our essentially wired AI system that picks these clothes for people now. The way it works under the hood, it's a little different from how others do it. I'm sure you heard of things like Stitch Fix, Trunk Club, and all these guys. The little difference is that, instead of having stylists being assisted by this AI to make these decisions, we do it the other way around. We have the AI make the decisions, and the stylists really look and see if they made these maybe improper decision making. The way it works, usually how people attack the problem or others attack the problem, is that each person they bring them up, and then look at the historical goals that they've… Because throughout the process when you signed up, you essentially are asked multiple questions. Those questions essentially help us guide our styling methods and questionnaires. The way we do it is we turn this process upside down. When we acquire a customer, we ask them a set of questions. Those questions could be from your sizing, your preference in patterns, the type of pants that you like. Then we go a little further, we look at your occupation, we look at your location, the location tells us a lot about you. That tells us the type of clothes that you want. If you send someone in Florida a heavy jacket, I'm sure they're not going to like it. The location makes a lot of sense. Occupation makes a lot of sense.