San Francisco (CNN Business)In 2010, artificial intelligence was more likely to pop up in dystopian science-fiction movies than in everyday life. And it certainly wasn't something people worried might take over their jobs in the near future. A lot has changed since then. AI is now used for everything from helping you take better smartphone photos and analyzing your personality in job interviews to letting you buy a sandwich without paying a cashier. It's also becoming increasingly common -- and controversial -- when used for surveillance, such as facial-recognition software, and for spreading misinformation, as with deepfake videos that purport to show a person doing or saying something they didn't.

Anything you do online can come back to bite you. It's been a decade full of lessons: who to trust, when to speak out and how to stream big events online after you've broken up with your cable company. In 2010, the first iPhone was only three years old. Uber and Lyft didn't exist, and neither did Google Assistant and Siri, Instagram or streaming video. We've come a long way since then, but the next 10 years won't be easy.

Banking communications and press releases show conversational interfaces accounting for 38.87% of the AI use-cases at banks. In truth, most chatbots are pilot projects with little to no evidence of ROI; they're touted in the press to make banks appear more modern and convenient to customers. Emerj's AI in Banking Vendor Scorecard and Capability Map found conversational interface vendors score lowest in terms of funding and the AI talent they employ (2.4 out of 4.0). The average customer service vendor in banking raises $16 million, far less than the average vendor in financing and loans ($49 million), fraud and cybersecurity ($48 million), and compliance ($44 million). Companies raise more money when their products have traction, and conversational interface vendors make up only 5.5% of the total funding for AI vendors.

If the idea of asking Alexa or Google to turn on and off your lights appeals to you, and you're not doing it already, the holidays could be a great time to finally get to it. "Competition is growing and prices are dropping which makes now the best time to make your home," smart, says YouTuber Steve Siems, who has a channel called "Steve Does," devoted to smart home reviews and installation. He suggests starting small, with a connected speaker, then adding smart switches and bulbs before venturing further with doorbells and other products. "See what you like and what you need more of," he says. "No need to buy 10 smart plugs then realize you only need three for what you want to do. By the time you use the other seven plugs, something newer, better, and cheaper will be out."

Google Assistant and Alexa may pretend to have "personality," but they really don't. Telling a joke when asked does not make any of them a great raconteur. This is fine for two reasons. First, it's not what they're for, and second, giving an artificial creation personality is very, very difficult. Gatebox, the company behind the eponymous product, is succeeding where others have either failed, or aren't even trying.

You may have to pay to speak to a real person when you agree to hybrid human-robo management. Technically, you are always in charge of your finances, but you may not be willing to hand over your portfolio's reigns to a robot. A robo-advisor may not be a great fit if you want a more hands-on approach to online guidance. Even an algorithm is still the most sophisticated computer algorithm. It can't sit with you, it can't explain anything to you, and it can't listen to your future dreams.

Recommender systems often rely on models which are trained to maximize accuracy in predicting user preferences. When the systems are deployed, these models determine the availability of content and information to different users. The gap between these objectives gives rise to a potential for unintended consequences, contributing to phenomena such as filter bubbles and polarization. In this work, we consider directly the information availability problem through the lens of user recourse. Using ideas of reachability, we propose a computationally efficient audit for top-$N$ linear recommender models. Furthermore, we describe the relationship between model complexity and the effort necessary for users to exert control over their recommendations. We use this insight to provide a novel perspective on the user cold-start problem. Finally, we demonstrate these concepts with an empirical investigation of a state-of-the-art model trained on a widely used movie ratings dataset.

The Talking Tech podcast is available for you every day with a quick hit on the latest tech news, gadget reviews, opinion on tech trends and interviews with insiders. On this page, you'll find quick links to all of our shows. Please rate and review the show on Apple Podcasts and "Favorite" us on Stitcher to help Talking Tech reach a wider audience. We have a new Alexa/Talking Tech skill! Follow USA TODAY's Jefferson Graham (@jeffersongraham) on Twitter, Instagram and YouTube.

The scene plays out like a thriller: You pull out your phone, and you see an ad for AirPods. Wait a minute, you think. Didn't I just have a conversation about AirPods with my friend? Like, a real conversation, spoken aloud? Is my phone… listening to me?

"Q: How worried do you think we humans should be that machines will take our jobs? A: It depends what role machine intelligence will play. Machine intelligence in some cases will be useful for solving problems, such as translation. But in other cases, such as in finance or medicine, it will replace people." This Q&A is taken from Tom Standage's description of how he interviewed AI (language model GPT-2) for The Economist The World in 2020. As readers of this column's annual roundup of AI predictions know, this year's first installment of 120 AI predictions for 2020 featured my interview of Amazon AI in which Alexa performed slightly better than the previous year. For the new list of 99 additional predictions, I repeated Standage's question to Alexa, and got the response "Hmm, I'm not sure." The following AI movers and shakers are a lot more confident in what the near future of machine intelligence will look like, from robotic process automation (RPA) to human intelligence augmentation (HIA) to natural language processing (NLP).