Google CEO-AI Needs Regulation; Netflix Happiest Workplace; Clearview Lets Strangers Find Your Info

Alphabet and Google CEO Sundar Pichai says AI is too important and potentially dangerous not to be regulated. According to 9to5google.com, he wrote an OpEd in the Financial Times, saying…among other things…’Now there is no question in my mind that artificial intelligence needs to be regulated. It is too important not to. The only question is how to approach it.’ Pichai added ‘Good regulatory frameworks will consider safety, explainability, fairness and accountability to ensure we develop the right tools in the right ways. Sensible regulation must also take a proportionate approach, balancing potential harms, especially in high-risk areas, with social opportunities.’

As if to add to his argument, a story out today covered by engadget.com says Kaylen Ward, the woman promising to send a naked pic of herself to donors to the Australian fires (and who has raised over a million dollars for fire relief doing so) got kicked off her 3 Instagram accounts…NONE of which were used for the fundraiser. Ward, ‘The Naked Philanthropist,’ had already ended her Twitter campaign for fire relief. Apparently, it’s not just Instagram that is the Morality Police. Airbnb is now using ‘trait analyzer’ algorithms that compile data dossiers on users, decide whether you’ve been bad or good, give you a score, and then “flag and investigate suspicious activity before it happens.”

The Evening Standard reported on Airbnb’s patent for AI that crawls and scrapes everything it can find on you, “including social media for traits such as ‘conscientiousness and openness’ against the usual credit and identity checks and what it describes as ‘secure third-party databases’.” They added, “Traits such as “neuroticism and involvement in crimes” and “narcissism, Machiavellianism, or psychopathy” are “perceived as untrustworthy.”

How about AI for job candidates? HireVue poses pre-determined questions, and analyzes words, grammar, facial expressions, and tone of the applicant’s voice to conclude ‘whether the candidate is tenacious, resilient, or good at working on a team, for instance.’ The problem with this is that humans can’t even do this…and who wrote the code? HUMANS! Psychiatrists and behavioral experts have warned that there is no way this AI analyzed interview can possibly accurately predict how the person will work on a team or if they are tenacious, and so forth. Beware!

On a happier note…unless you are the House of Mouse…the so-called ‘Happiest Place on Earth,’ Netflix has the happiest US employees for 2019. Businessinsider.com reports that this is the result from anonymous professional networking app Blind. Almost 86% of Netflix employees said they are happy at work. #2 and trying harder…Bloomberg. Netflix got the top spot despite its rep for being a high performance culture that is quite demanding, and that routinely fires underperformers. If you are applying there, it apparently helps for a present employee to have recommended you to the recruiters…like numerous other businesses across the country and world! Sorry, Disney…you will have to try harder to be the Happiest Place next time.

In more creepy AI news, startup Clearview uses a photo from its database of some 3 billion pictures scraped from Facebook, Venmo, YouTube, and other sites, and claims its AI can match up names and other info with those photos. Cnet.com says the database is now being used by hundreds of law enforcement agencies, as it is an order of magnitude bigger than that of the FBI, which claims 641 million images. The database isn’t available to the public right now, but could be opened to businesses and the public.

The House Oversight Committee is looking into this, and has already held 3 hearings on this type of use of AI. In November, 2 senators introduced a bill that would limit the use of this type of tech by law enforcement including the FBI. ICE, and others. “Facial recognition technology can be a powerful tool for law enforcement officials,” one of the senators, Mike Lee, a Republican from Utah, said in a statement at the time. “But its very power also makes it ripe for abuse.”


Apple Gears Up v DOJ Over Unlocking; Dorsey-Twitter Won’t Get Edit; Facebook- Notify re 3rd Party Logins; Lucid Readies Production of Air EV

Now that the president has joined the Department of Justice in demanding Apple to help the DOJ unlock (hack, really) the phones used by the Pensacola shooter, Cupertino is gearing up for a legal battle. Businessinsider.com reports that Apple has always refused to build a back door for governments, and it will continue to do so. Apple claims they are working with authorities to assist in the case, but has said the DOJ hasn’t spent enough time using third party tools to try to unlock the phones. In a previous case in 2016, the government and Apple got to the brink of this type battle over the San Bernardino shootings, but that battle went away when a third party vendor came forward with a way to unlock the killer’s phone. Apple has always been concerned that a back door would be hackable by others in and out of government, and would be too big a compromise of its users’ privacy.

During a Q&A with Wired, Twitter CEO Jack Dorsey says that Twitter will likely not be adding an edit button for users to clean up their Tweets. According to engadget.com, Dorsey says he has gone back and forth on this himself, but said that ultimately Twitter had decided to preserve the SMS type feel, and so will keep it like your texts…that is, you can’t take ‘em back once sent. Additionally, Dorsey pointed out that if people re-Tweet your Tweets, then you edit it, the original is what was re-Tweeted, and it could differ greatly from what you ultimately ended up saying.

With so many sites that offer logins via your Facebook credentials, you can easily lose track of where you have used that method. Now, Facebook will help, by sending you a notification when your account is used to log into a third party app. Zdnet.com says the social net will send you an email, and it will give you info about the details Facebook shared with the third party app to log in. If you decide not to let an app keep using your Facebook login, you an go to Edit Settings, and revoke that app’s access permissions.

Lucid Motors plans to roll out the production version of its Air EV in April in New York. According to electrek.co, they will price the electric sedan at $60,000 for the base price, with a range of 240 miles. The factory is being built in Casa Grande, Arizona, but Lucid is building out 80 beta prototypes at its California HQ for testing. They are shooting for full production by late 2020. Lucid claims to be working on a ‘network’ of retail and service locations.


Playstation 5 Alleged Drop Date & Price; Google Buys Pointy; RIP Win 7 & Major Win Bug; Amazon AI- Why People Buy Irrelevant Stuff

Sony showed off the Playstation 5 logo last week at CES. Now, thanks to a big leak, it looks like we have the release date and pricing info. Bgr.com reports that the leak has the PS5 bowing at a PlayStation meeting on February 5th at Sony Hall in New York City. It’s expected that the console design, controller, home screen, specs, and of course games will be shown. The rumor touts ‘little to no load times,’ ‘blazing fast downloads,’ ‘immersive controls,’ and claims Sony will use ‘It’s time to play’ as their new PS5 slogan…and may apply that tag line to the entire brand. What about the pricing?? $499 is the amount leaked. There will apparently only be a single model out at launch, and no Pro version will make it into stores in 2020. It will supposedly have 10 teraflops of GPU power, compared to 4 of the Xbox. Another rumor has a more expensive Xbox sporting 12, however. The leak claims Sony will offer full backwards compatibility with all PS4 games, which is huge. Also…saved data/ backups form PS4 games will be transferable.

Google seems to be moving in on some of Amazon’s territory, picking up Pointy…a startup that helps brick and mortar retailers to list products online (I’m just smelling a whiff of Amazon Marketplace here). According to techcrunch.com, Pointy has developed hardware and software tech to help retailers…especially those that don’t already have deep e-commerce presence…to get their inventory findable online without any extra work. Pointy is a startup out of Dublin, Ireland. It looks like for now, Google will let the company operate relatively independently. Google has partnered with them since 2018. The original Pointy hardware is only $700, and they have a free app that integrates with point of sale equipment from the likes of Clover, Square, Lightspeed, Vend, Liberty, WooPOS, BestRx, and CashRx POS, so a retailer can essentially get into their ecosystem for nothing at first. It seems pretty likely that Google will be looking to scale this up into a major presence.

You’ve probably heard the news that today is the day Microsoft ends support for venerable Windows 7. Ironically….or maybe not…security expert Brian Krebs warns that there is a nasty bug that affects not only Windows 7, but Vista as well. Pcworld.com says Microsoft may be pushing out a patch for it today. The bug allegedly concerned crypt32.dll, which controls ‘certificate and cryptographic messaging functions in the CryptoAPI.’ Infection creates the possibility of your PC allowing a piece of malware to be installed posing as a legit application.

Amazon is using AI to try to figure out why people buy products that seem irrelevant to their searches (both web and voice ones.) According venturebeat.com, the AI group has found that people are most likely to do so in a few categories…like toys and digital products…and not so much in beauty products or groceries. An interesting result so far…it turns out that people who use either very short or unusually long queries tend to be more ‘flexible’ in their purchasing decisions than people who use medium length ones. So if you go for the really short ones or long ones, you’re apparently a soft touch!


More Galaxy Leaks; 5G iPhones Set; GM May Roll Out e-Hummer; Some Walmarts Getting Inventory Robots

As we get closer to Samsung’s Unpacked event on February 11th, more keeps coming out about the next gen phones. 9to5google.com reports that the handsets will definitely be named Galaxy S20, skipping ahead a decade in numbering. Also, images leaked out over the weekend, showing the expected ‘domino bump’ on the back with the 4 cams, flash, etc. The phones won’t have the curved glass as previous models, and have a very slight bezel. The top line Galaxy S20 Ultra will be the top line, but the #2 phone…S20+ …will be the one to sport a new 12MP main cam, in addition to an ultra wide angle and telephoto zoom lens. The selfie cam will be a hole punch type, so the phones will have very slight ‘forehead’ and ‘chin’ expansions of the bezel at the top and bottom.

Noted Apple watcher and analyst Ming-Chi Kuo has now said he sees 4 5G phones coming from Apple later this year, including both sub-6GHx and mmWave versions…depending on the part of the world. According to macrumors.com, the mmWave phones will be limited to the US, Canada, Japan, Korea, and the UK. Kuo thinks the phones will launch in the US in September as usual, but some sources say they won’t make it out until late in the 4th quarter. At any rate, don’t expect to have access to 5G service widely in 2020…the providers are building out as fast as possible, but like with LTE when it first bowed, expect your phone display to jump between 5G and LTE (or 5Ge, as AT&T has tried to re-brand it) for at least a year or two!

After killing the gas-guzzling, excess-dripping brand following their bankruptcy in 2009, GM is toying with reviving the Hummer brand. Mashable.com says the new Hummers would be all electric…starting with an e-SUV in 2022. They also are planning a full sized e-pickup. Since GM already has a number of mid-sized and small EVs in the pipeline, it’s expected that the e-Hummers will play off the original’s brand, and be full sized EVs. The new Hummer pickup is slated to be shown off by LeBron James in a Super Bowl ad next month.

While others are grappling with robots that have arms and can restock shelves, Bossa Nova is getting ready to launch a battalion of inventory robots at 1,000 US Walmart stores. According to techcrunch.com, the devices, which look like a roller suitcase with a tall, stationary raised ‘arm’, will only scan the shelves for inventory, and transmit back the info to a central computer. This will enable good ol’ humans to go out and restock the shelves with precisely what’s needed. Bay Area based Bossa Nova had previously done a test run with 350 of the devices. The rollout of 1,000 inventory bots should be in place by summer…covering about 20% of Walmart’s 5,000 US stores.


Apple’s App Store- Record 2019; Facebook & eBay Step Up Fake Review Policing; Ford Getting 2 Leg Robot for Package Delivery; BMW Shows Future AR Windshield

Apple says its App Store brought in a record $1.42 billion for developers during the Christmas Eve to New Year’s Eve period…a 16% jump over last year. The amount included app sales and sales of in-app items. Cnet.com reports that on New Year’s Evey Day, Apple hit a single-day record of $386 million, a 20% boost over last year! Which apps did the best?

Top Free iPhone Apps
1 YouTube: Watch, Listen, Stream
2 Instagram
3 Snapchat
4 TikTok – Make Your Day
5 Messenger
6 Gmail – Email by Google
7 Netflix
8 Facebook
9 Google Maps – Transit & Food
10 Amazon – Shopping made easy
Top Paid iPhone Apps
1 Facetune
2 HotSchedules
3 Dark Sky Weather
4 The Wonder Weeks
5 AutoSleep Tracker for Watch
6 TouchRetouch
7 Procreate Pocket
8 Sky Guide
9 Toca Hair Salon 3
10 Scanner Pro: PDF Scanner App

Under pressure from the UK, Facebook and eBay have promised to do more to clean up fake reviews. According to techcrunch.com, a British consumer rights group says Facebook has removed 188 groups and 24 user accounts so far. eBay has permanently banned 140 users after the watchdog and UK’s Competition and Markets Authority pressed them. Both companies have pledged to ‘better identify, investigate and respond to’ the trade in fake reviews, and keep them off their platforms moving forward.

Ford has announced the purchase of some new two-legged robots built by Agility Robotics. Zdnet.com says the robot, named Digit, is a 5 foot tall robot with humanoid form factor. Ford plans to mate them up with self-driving delivery vehicles to handle the pesky ‘last 50 feet’ of delivery that has been a knotty problem for delivery services. Some are working on drones…like Amazon, but those have limited carrying capacity and range. Ford would like to have autonomous vehicles for commercial operation by next year. Unlike a small drone, the robot could unfold out of the back of the vehicle, pick up a package of up to 40 lbs, and put it on a doorstep. Since Digit has legs as opposed to wheels like other concept delivery robots, stairs aren’t a problem. Ford plans for the robots to be able to top of and recharge while riding from place to place in the vehicle. Lidar will spot potential obstacles for the robot, and it can radio back to the car, which can connect to a bigger computer that will send a solution. It’s doubtful this will all happen by 2021 as Ford might like, but it’s definitely on the way.

In other cool auto tech, BMW showed off a mind blowing concept…an AR windshield for the 2021 iNext. According to engadget.com, the windshield has an overlaid transparent display, that lets data be overlaid onto the road ahead. You might get descriptions and contextual info about nearby objects, buildings, and signs. It has eye tracking sensors that let the car follow your gaze, and display more data about something that has caught your attention. If you want to go deeper, you will be able to gesture towards the screen, and more info will appear. To be able to make all this happen, it will require the speed of 5G, and BMW says the 2021 iNext will be the first car to feature 5G standard.


Facebook Bans Deepfakes-But Not Other Fake Videos; Galaxy Chromebook Bows; Lenovo ThinkBook Plus; Waymo-20 Million Public Road Miles

The good news: Facebook has banned deepfakes. The bad news…they won’t ban other fake videos, NOT made with the AI used to make deepfakes. 9to5mac.com reports that the platform will at least spare us seeing the super realistic fake videos that can analyze how peoples’ faces move when they talk and mimic the movement needed for any word. Deepfakes can create fake videos from chopped up sound clips of an actual person or utilize an impressionist to create videos that are hard to tell from actual video made of the person. The tech has existed at least since 2017, when University of Washington used to to create a fake Obama speech. Because only the Deepfake AI tech videos are banned, we will still be bombarded with videos and pictures made with normal editing tools…like the fake video doctored to make Speaker Pelosi appear to slur her words that appeared earlier. Maybe more than politicians…deepfakes really freak out actors and actresses…which can be made to appear in fake porn videos doing things they would never do…at least in front of a camera.

Samsung rolled out a Galaxy Chromebook at CES. It’s a great looking machine, but at $999 starting price, pretty sky high for a Chromebook. It does come with a 13.3 inch AMOLED display, and a 10th gen Intel Core i5 processor. According to theverge.com, there’s a fingerprint sensor for unlocking, a pair of USB-C ports, and expandable storage via micro-USB. The screen rotates 360 degrees, and an S-Pen is included, which has a handy storage slot. At least for the almost thousand bucks, it’s aluminum, and not plastic. Besides the price point, this box won’t be for everyone….it comes in a bright red orange!

Lenovo showed off a number of new laptops at CES, but one was a bit of a surprise. Engadget.com notes that it’s the ThinkBook Plus. It comes with a 13.3 inch matte full HD IPS display, a single USB-C Gen 2 port, a couple standard USB ports, and and HDMI out. the rig can be spec-ed out with up to a 10th ten Core i7 processor, up to 16 gigs of RAM, and 512 gigs of storage. That’s all cool, but on top of it, there’s a 10.8 inch touch sensitive E-Ink display on the back of the screen! If you want to check an email or your calendar, you can do it without opening up your device at all (although you can’t edit the calendar or reply to email.) You can use the 2nd display to jot down notes with a stylus in OneNote. You can also talk to Alexa over the E-Ink screen. The demo version had to be touched to activate it…Lenovo says on release, you will be able to talk to Alexa as you normally would. No pricing was given at CES.

Google self-driving car division Waymo has announced that its cars have now driven over 20 million miles over public highways, and been run through tens of billions of miles in computer simulations. According to venturebeat.com, the roadway driving has mainly taken place in Novi, Michigan, Kirkland, Washington, and San Francisco. The public miles is double what Waymo had racked up just a year ago! For comparison, Yandex and Baidu both claim to have driven about a million miles each, and GM’s Cruise division says it’s on the verge of hitting a million miles a month with their autonomous test cars. Not everyone is impressed…Uber claims miles driven doesn’t mean much without context like location…a sentiment shared by the Department of Transportation. Way did pass 1500 monthly active riders in December, and has now hauled over 100,000 since launching its programs in 2017.


Samsung Unpacked Date & Rumors; Apple- 9 of Top 10 Christmastime Activations; CES Insults Own Cannabis Award Winner; Illinois Bans AI Graded Hire Videos

Samsung Unpacked will be February 11th in San Francisco’s Palace of Fine Arts. Rumors have had the handsets named Galaxy S20, but until it’s known for sure, we’ll say 11 here. There are expected to be 3 S11 phones, and a foldable one with the form factor of Motorola’s Razr, not the size and shape of the Galaxy Fold. Cnet.com notes that Samsung has rolled out its phones before at Mobile World Congress in Barcelona, which is coming up February 24th, but this gives them a head start on rivals. The new generation Galaxy phones are rumored to have 5G, a 108!! megapixel camera, 5,000 may battery, and a new form factor that resembles the Galaxy Note 10.

Apple made nearly a clean sweep of device activations for the holiday period of December 25th-31st. According to Flurry Analytics, Cupertino took 9 of the top ten spots, led by the iPhone 11. Older models…the XR, iPhone 7, and iPhone 8…took the next 3 slots, with the 11 Pro Max coming in 5th. The only non-Apple phone in the top 10 was the Xiaomi Mi 4 LTE at #10. Over the holiday period, Apple bagged 42.8% of all mobile device activations when you include iPads. Samsung held 23.5% of the market through the holidays, leaving everyone else scrapping over the remaining 33.8%

Shades of last year…CES this year awarded Keep Labs, giving them Innovation Award Honoree, but won’t let them say or display the word ‘cannabis’ on the CES show floor! Keep Labs has decided to pass on exhibiting due to having their marquee product basically banned. It is still listed on the CES website. According to techcrunch.com, last year, CES banned Honoree Innovation Award winner Lora DiCarlo, which makes sex toys. CES also withdrew the award, but after quite an uproar, restored the award and apologized. Lora DiCarlo successfully lobbied, arguing that the CES was discriminating against a product for women…and this year they will be on the show floor with 2 new products that use micro-robotics tech to mimic human touch. One gadget, Onda, is G-spot specific, and the other named Baci is designed to target the clitoris. The original Ose claims to produce a ‘blended orgasm.’ NO, you can’t try before you buy on the show floor!

As computer tech becomes more and more pervasive in hiring…software has screened resumes for a couple decades now…Illinois has moved to block a type of AI in hiring use… due to possible bias and inaccuracy. Vox.com says the type of software used by HireVue uses algorithm based analysis of recorded video interviews. The algorithm analyzes how you answered AND your facial expressions to make predictions about your fit for a position. The Illinois lawmakers considered that most facial recognition software is only 85-95% accurate, and restricted the AI. The law requires firms to explain how the AI works, and wheat ‘general types of characteristics’ it considers when evaluating candidates. The law also limits who can view the recorded video interview to those ‘who’s expertise or technology is necessary’ and requires that companies delete any video an applicant submits within a month of their request. It’s not exactly a heavy handed law, but does give applicants at least some rights when computers are involved in trying to get a job and make a living. Welcome to the future, and all hail our computer and robotic overlords….