Daily Tech Snippet: Thursday, February 2nd

• Facebook eases past Wall Street estimates, sees spending up in 2017: Facebook Inc (FB.O) cruised past Wall Street's earnings and revenue expectations on Wednesday with strong growth in its mobile ad business, demonstrating that controversy over so-called "fake news" and inaccurate advertising measurements had little impact on its financial performance. With quarterly profit of $3.57 billion, more than double the $1.56 billion it reported a year ago, the company showed no signs of slowdown in growth. The results handily beat analysts' expectations, and shares ticked up about 0.2 percent in after-hours trading. The company had warned in November that ad growth would likely slow "meaningfully" due to limits on ad load - the total number of ads Facebook can show to each user. But there was little sign of that in the fourth quarter as total revenue soared to $8.81 billion from $5.84 billion a year ago.

• Facebook Loses $500 Million Virtual Reality Headset Verdict: The virtual reality headset maker that Facebook Inc. bought in 2014 for $2 billion used stolen computer code, a jury said in awarding $500 million to ZeniMax Media Inc. The case was over the Oculus Rift, the device that has put the social media giant at the forefront of the virtual reality boom. Wednesday’s verdict in Dallas federal court is a rebuke to Facebook Chief Executive Officer Mark Zuckerberg, who isn’t a defendant but who told jurors in his first-ever courtroom testimony that it was important for him to be there because the claims by ZeniMax were “false.” An Oculus spokeswoman said the company will appeal. The case centered around the defection of video-game programmer John Carmack from ZeniMax, where he had designed blockbuster games Doom and Quake, to Oculus, where he was named chief technology officer in 2013. He acknowledged in testimony that he took with him e-mail records including computer code related to virtual reality.

• H&R Block is now using IBM Watson to find tax deductions: Getting your taxes done is miserable. It’s complicated, time-consuming and costly — particularly for people whose financial situation requires them to seek help from tax professionals. The process is such a mess, even for said professionals, that H&R Block, one of the largest tax services providers, is partnering with IBM Watson to help make the process go more smoothly. Yes, that’s right — we now need machine intelligence to save us from our taxes. Watson will use natural language processing to parse tax documents and relate statements back to the 74,000-page federal tax code. It’s not hard to see how a machine would excel at this clearly defined search task. These are just the sorts of problems that machines tend to outperform humans in solving. There are a number of things that could be automated in the tax preparation process, but H&R Block decided to focus on identifying deductions and credits. The consumer-facing approach will do the firm well. At the end of the day, most people really only care about paying as little in taxes as possible and maximizing the size of their refund. Using tax documents as a corpus for training Watson, IBM was able to fine-tune its system for H&R Block. Because Watson is constantly receiving more tax documents year in and year out, it will improve over time. H&R Block will benefit from a more accurate artificial companion each new tax season. This is on top of any changes IBM makes to the core Watson infrastructure.