The Xbox One is finally adding keyboard and mouse support for gamers who prefer to play that way. This is happening more than two years since Microsoft first suggested keyboard and mouse support was on its way. So, this is definitely a case of better late than never.
Xbox One Keyboard and Mouse Support
Microsoft officially announced it was adding Xbox One keyboard and mouse support in September. Since then, selected gamers have been testing it out. And now, as part of XO18, Microsoft has detailed which games are getting mouse and keyboard support.
ICYMI, Larry and Jeff just dropped a ton of mouse & keyboard news during the #X018 episode of Inside Xbox. Here is a quick recap:
1. 14 additional games confirmed to support mouse and keyboard input – including Fortnite – some starting as soon as next week. pic.twitter.com/P9kPVdYCa7
Fortnite and Warframe are the first two Xbox One games to support keyboard and mouse, and both should receive updates on November 14. These are both free-to-play shooters, with Fortnite being the game that all kids love and most parents don’t quite understand.
Bomber Crew, Strange Brigade, Warhammer: Vermintide 2, War Thunder, X-Morph: Defense, and Deep Rock Galactic will add support in November. And Children of Morta, Vigor, Warface, Wargroove, DayZ, Minion Masters, and Moonlighter will follow in the future.
With mouse and keyboard support having arrived on the Xbox One, Microsoft is hoping to take advantage of the situation. The company has partnered with Razer to develop an official line of Xbox keyboards and mice. The first of which will be unveiled at CES 2019.
Keyboard and Mouse Support for More Games
PC gamers who switch to consoles often have trouble adapting to a controller. So, given the crossover between PC and Xbox One games, this makes perfect sense for Microsoft. Now it’s up to developers to implement mouse and keyboard support into more games.
Cloudflare has launched a mobile app for its 1.1.1.1 DNS service. This makes it easier for smartphone users (on Android and iOS) to use. Cloudflare’s DNS service, which is designed to improve privacy and performance, can also be used on your desktop.
What Is Cloudflare’s 1.1.1.1 DNS Service?
In April 2018, Cloudflare launched its DNS service, catchily monikered 1.1.1.1. This is a DNS resolver that anyone can use for free. We previously detailed how to set it up on Windows and Mac, and explained how 1.1.1.1 solves some big privacy issues.
Using Cloudflare’s 1.1.1.1 DNS service should help you keep your browsing habits private from your ISP and other interested parties. It should also help improve performance, allowing you to load websites faster than you can using your ISP’s default DNS server.
Cloudflare’s 1.1.1.1 DNS Lands on Mobile
After a successful beta, Cloudflare has launched its 1.1.1.1 DNS service as a mobile app for Android and iOS. Cloudflare describes 1.1.1.1 as “the world’s fastest and safest DNS resolver” which is focused on a combination of privacy and performance.
Once you install 1.1.1.1 you can turn it on or off with just a few clicks. When it’s turned on your internet usage will be funnelled through a secure connection. This makes sure no one can snoop on your DNS queries, whether it’s your ISP or the government.
By making use of Cloudflare’s global network, the company claims 1.1.1.1 can improve your internet performance. Either way, 1.1.1.1 is free (for life), and Cloudflare promises to “never sell your data or use it to target ads”. Unlike some ISPs.
Download: Cloudflare’s 1.1.1.1 on Android | iOS (Free)
Microsoft has slowly but surely pushed the command line aside in the Windows interface. This is not without reason; it is an antiquated and mostly unnecessary tool from an era of text-based input. But many commands remain useful, and Windows 8 and 10 even added new features.
Here we present the essential commands every Windows user needs to know.
In case you’re not sure how to access the Windows command prompt, forgot basic Windows commands, or would like to know how to see a list of switches for each command (aka prompt codes), you can refer to our beginners guide to the Windows command line for instructions.
Prefer this tutorial in video form? We’ve got you covered:
Most files in Windows are associated with a specific program that is assigned to open the file by default. At times, remembering these associations can become confusing. You can remind yourself by entering the command “assoc” to display a full list of file name extensions and program associations.
You can also extend the command to change file associations. For example, “assoc .txt=” will change the file association for text files to whatever program you enter after the equal sign. The “Assoc” command itself will reveal both the extension names and program names, which will help you properly use this command. You can probably do this more easily in the GUI, but the command line interface is a perfectly functional alternative.
Deleting files on a mechanical hard drive doesn’t really delete them at all. Instead, it marks the files as no longer accessible and the space they took up as free. The files remain recoverable until the system overwrites them with new data, which can take some time.
The cipher command, however, wipes a directory by writing random data to it. To wipe your C drive, for example, you’d use the command “cipher /w:c”, which will wipe free space on the drive. The command does not overwrite undeleted data, so you will not wipe out files you need by running this command.
You can use a host of other cipher commands, however, they are generally redundant with BitLocker enabled versions of Windows.
Drivers remain among the most important software installed on a PC. Improperly configured or missing drivers can cause all sorts of trouble, so its good to have access to a list of what’s on your PC. That’s exactly what the “driverquery” command does. You can extend it to “driverquery -v” to obtain more information, including the directory in which the driver is installed.
You can use this command to identify differences in text between two files. It’s particularly useful for writers and programmers trying to find small changes between two versions of a file. Simply type “fc” and then the directory path and file name of the two files you want to compare.
You can also extend the command in several ways. Typing “/b” compares only binary output, “/c” disregards the case of text in the comparison, and “/l” only compares ASCII text.
So, for example, you could use the following:
fc /l "C:\Program Files (x86)\example1.doc" "C:\Program Files (x86)\example2.doc"
The above command compares ASCII text in two word documents.
This command relays the IP address that your computer is currently using. However, if you’re behind a router (like most computers today), you’ll instead receive the local network address of the router.
Still, ipconfig is useful because of its extensions. “ipconfig /release” followed by “ipconfig /renew” can force your Windows PC into asking for a new IP address, which is useful if your computer claims one isn’t available. You can also use “ipconfig /flushdns” to refresh your DNS address. These commands are great if the Windows network troubleshooter chokes, which does happen on occasion.
Entering the command “netstat -an” will provide you with a list of currently open ports and related IP addresses. This command will also tell you what state the port is in – listening, established or closed.
This is a great command for when you’re trying to troubleshoot devices connected to your PC or when you fear a Trojan infected your system and you’re trying to locate a malicious connection.
Sometimes, you need to know whether or not packets are making it to a specific networked device. That’s where ping comes in handy.
Typing “ping” followed by an IP address or web domain will send a series of test packets to the specified address. If they arrive and are returned, you know the device is capable of communicating with your PC; if it fails, you know that there’s something blocking communication between the device and your computer. This can help you decide if the root of the issue is an improper configuration or a failure of network hardware.
This is a more advanced version of ping that’s useful if there are multiple routers between your PC and the device you’re testing. Like ping, you use this command by typing “pathping” followed by the IP address, but unlike ping, pathping also relays some information about the route the test packets take.
The “tracert” command is similar to pathping. Once again, type “tracert” followed by the IP address or domain you’d like to trace. You’ll receive information about each step in the route between your PC and the target. Unlike pathping, however, tracert also tracks how much time (in milliseconds) each hop between servers or devices takes.
Powercfg is a very powerful command for managing and tracking how your computer uses energy. You can use the command “powercfg hibernate on” and “powercfg hibernate off” to manage hibernation, and you can also use the command “powercfg /a” to view the power-saving states currently available on your PC.
Another useful command is “powercfg /devicequery s1_supported”, which displays a list of devices on your computer that support connected standby. When enabled, you can use these devices to bring your computer out of standby — even remotely. You can enable this by selecting the device in Device Manager, opening its properties, going to the Power Management tab and then checking the Allow this device to wake the computer box.
“Powercfg /lastwake” will show you what device last woke your PC from a sleep state. You can use this command to troubleshoot your PC if it seems to wake from sleep at random.
You can use the “powercfg /energy” command to build a detailed power consumption report for your PC. The report saves to the directory indicated after the command finishes. This report will let you know of any system faults that might increase power consumption, like devices blocking certain sleep modes, or poorly configured to respond to your power management settings.
Windows 8 added “powercfg /batteryreport”, which provides a detailed analysis of battery use, if applicable. Normally output to your Windows user directory, the report provides details about the time and length of charge and discharge cycles, lifetime average battery life, and estimated battery capacity.
This is, of course, redundant with the already easily accessed shutdown button, but what’s not redundant is the “shutdown /r /o” command, which restarts your PC and launches the Advanced Start Options menu, which is where you can access Safe Mode and Windows recovery utilities. This is useful if you want to restart your computer for troubleshooting purposes.
This command will give you a detailed configuration overview of your computer. The list covers your operating system and hardware. For example, you can look up the original Windows installation date, the last boot time, your BIOS version, total and available memory, installed hotfixes, network card configurations, and more.
Use “systeminfo /s” followed by the host name of a computer on your local network, to remotely grab the information for that system. This may require additional syntax elements for the domain, user name, and password, like this: “systeminfo /s [host_name] /u [domain]\[user_name] /p [user_password]”
You will need to run the command prompt with administrator privileges and enter the command “sfc /scannow”. If SFC finds any corrupt or missing files, it will automatically replace them using cached copies kept by Windows for this purpose alone. The command can require a half-hour to run on older notebooks.
You can use the “tasklist” command to provide a current list of all tasks running on your PC. Though somewhat redundant with Task Manager, the command may sometimes find tasks hidden from view in that utility.
There’s also a wide range of modifiers. “Tasklist -svc” shows services related to each task, use “tasklist -v” to obtain more detail on each task, and “tasklist -m” will locate .dll files associated with active tasks. These commands are useful for advanced troubleshooting.
Tasks that appear in the “tasklist” command will have an executable and process ID (a four- or five-digit number) associated with them. You can force stop a program using “taskkill -im” followed by the executable’s name, or “taskkill -pid” followed by the process ID. Again, this is a bit redundant with Task Manager, but you can use it to kill otherwise unresponsive or hidden programs.
Virtually all Windows 8/8.1 computers ship from the factory with a recovery image, but the image may include bloatware you’d rather not have re-installed. Once you’ve un-installed the software you can create a new image using the “recimg” command. Entering this command presents a very detailed explanation of how to use it.
You must have administrator privileges to use the “recimg” command, and you can only access the custom recovery image you create via the Windows 8 “refresh” feature.
In Windows 10, system recovery has changed. Windows 10 systems don’t come with a recovery partition, which makes it more important than ever to back up your data.
CMD Commands Recommended by Readers
These are the commands our readers use regularly:
nbstat: “For looking up names of computers on your network.” (Vferg)
netstat -ano | find “est”: “To get a list of processes with established CP connections.” (Eric)
tasklist | find “[process id]”: “To get the name of the executable associated with the particular process id that I’m interested in.” (Eric)
cacls: This command is “most handy to manually access hidden files and folder. (A41202813)
net use: “To map drives of networked CNC machines.” (Jimbo)
chkdsk /f C: “Checks your C: partition hard disk for errors and fixes bad sectors.” (Kai M)
Schtasks: To schedule tasks. (Teddy)
Command and Conquer Your Windows PC
This article can only give you a taste of what’s hidden within the Windows command line. When including all variables, there are literally hundreds of commands. Download Microsoft’s command line reference guide (in Edge or Internet Explorer) for advanced support and troubleshooting.
With Windows 10, Microsoft has moved from the command prompt to PowerShell. While you can still access the Windows command line, it may be time to switch. Note that you can use command line commands in the Windows PowerShell.
You just cannot create beautiful Microsoft Word documents by cutting corners on tables. On Microsoft Word, tables are essential formatting tools. Microsoft Office has made it easier to create and format basic tables in Word for Office 365, Word 2019, Word 2016, andWord 2013.
We haven’t covered tables as much as we would have liked to. It’s time to correct that as the number of people asking questions on how to format tables properly is piling up. Maybe these eight table tips can be an appetizer.
1. How to Make a Table in Word
Using tables and even changing them on the fly according to the data has become far easier in the newer versions of Word like Microsoft Word 2019 and Office 365. Intuitive formatting features give you finer (and quicker) control over how a table looks. But first head to the Ribbon > Insert > Table > Insert Table for making your first table.
It gives you five options for creating your first table.
The quickest way to start is with Quick Tables. The built-in designs save you from the lack of design skills. You can modify the designs by adding your own rows and columns or deleting the ones you don’t need.
Another quick way to create a table in Word is the Insert Control feature. You can create a new column or row with one click. Hover the mouse over a table. A bar appears right outside your table between two existing columns or rows. click on it when it appears, and a new column or row will be inserted at that position.
When you want to move or order a row around, use the combination of Alt+Shift+Up Arrow and Alt+Shift+Down Arrow to order the row up or down. Move contiguous rows by selecting them all first.
How to Position Your Table on the Page?
Right-click on the table and select Table Properties from the context menu. The Table Properties dialog box is for precise control over the data and its display. Control the size, alignment, and indentation of the table.
By default, Word aligns a table on the left. If you want to center a table on the page, select the Table tab. Click on Alignment > Center.
The Indent from left figure controls the distance of the table from the left margin.
Position the table according to the text around it for a visually aesthetic look. Wrap text around tables by dragging it by the handle. The text wrapping changes automatically from None to Around. From the Table Positioning dialog box, you can set the Distance from surrounding text for each side of the table.
Select Move with Text if the text is directly related to the table data. The table is vertically aligned to the related paragraph around it. If the table data applies to the whole document, you can keep the option unchecked.
2. Use the Ruler
Sizing tables and positioning them accurately is an art in itself. If you need precise measurements to size your rows and columns—use the ruler.
Hover the mouse over a border. When the double-arrow pointer appears, click the border and hold down the ALT key. Microsoft Word shows specific measurements on the ruler. Move the rows and columns to fit your measurements.
3. Convert Text to Table (and Vice Versa)
Tabular data gives information its structure. It would have been frustrating if Word didn’t have something to handle non-tabular data. You can convert data to tables instantly from the Insert Table command.
Select the text. Go to Ribbon > Insert > Table > Insert Table
Microsoft Word determines the required number of rows and columns by considering the text separators and then auto-fits the contents. The Convert Text to Table dialog box allows you more control if the previous operation doesn’t work out right. You can also choose how to fit the contents of the table on the page.
You can specify how Microsoft Word should separate the data into rows and columns. Paragraph, tabs, commas, or any other delimiting character. This allows you to easily import non-tabular data from CSV files or plain TXT files and convert them into formatted tables.
Convert Table to Text
Engineer the reverse process if someone asks you to send them files with comma separated values or any other delineator. Select the entire table by clicking the “move” handle above the table.
Go to Ribbon > Table Tools > Layout > In the Data Group, click Convert to Text.
Simple text can be boring. When you have the chance, convert your table of data to a more visual chart instead with one of the underused features in Microsoft Word.
4. Auto-Fill Column Numbers
Microsoft Excel makes auto-filling a sequence of numbers very easy. Microsoft Word does not and you may have to resort to a manual job. There is a simpler way.
Create a new column for the serial numbers if it does not exist. Select this column by positioning the mouse over the column.
With the column selected, go to Home > Paragraph > Click the Numbering button for inserting a numbered list.
A number sequence is inserted in the column automatically.
5. Freeze Those Tables!
Microsoft Word tables change their dimension to accommodate new data. There may be times when you do not want the table to change size at all, even when new data is inserted. That is—“freeze” the size of cells.
The first step is to specify a fixed size for the cells. Go to Table Properties > Row > Enter a value in the Specify height box. For Row height is select Exactly from the dropdown.
Now, select the Table tab > click the Options button > uncheck the Automatically Resize to Fit Contents check box.
Click OK twice to exit the Table Properties dialog box.
This also solves the problem of inserting an image into a cell without the cell expanding to accommodate the image. If the image is bigger than the available space in the cell, it gets cropped to fit within the cell.
6. Change Rows Into Columns in a Table
There are situations where you have to change rows into columns and columns into rows. One possible scenario is where the number of columns exceeds the page margin. Switching columns around to rows and vice-versa is called transposition.
The bad news is that Word does not have an inbuilt method for handling this yet. Microsoft suggests that you copy-paste your table into Microsoft Excel and use its Transpose command. The transposed table can now be copy-pasted back into Microsoft Word.
You will find a use for this simple workaround. By default, Gmail does not retain the spreadsheet format when you paste from Microsoft Excel. To email tabular data without sending it as a separate attachment, use Microsoft Word as a bridge.
Select and copy-paste the Microsoft Excel table to a Microsoft Word document with the source formatting. Now, copy-paste from Microsoft Word to Gmail. As you can see from the screenshot, the problem is solved. You might have to tweak the more heavily formatted tables slightly, but most of the formatting is retained.
8. Reuse Your Tables to Save Time
You can save a lot of time by re-using tables when you create professional reports and documents. Save empty table formats and insert new data when required. With this quick save, you won’t have to recreate the layout from scratch for new data.
Select a table. Go to Ribbon > Insert > Text group > click Quick Parts > Save Selection to Quick Part Gallery.
After you save a selection to the Quick Part Gallery, you can reuse the selection by clicking Quick Parts and choosing the selection from the gallery.
Use the Building Blocks Organizer to preview any table you created. You can also edit properties and delete the tables from here.
Do Microsoft Word Tables Confuse You?
These tips aren’t enough to cover the scope of tables in our everyday work. I haven’t talked about the role of the Design tab in creating eye-catching tables. That is a topic in itself. But it is one of the lesser areas to get confused over thanks to the visual help in that tab.
Tables are one common area between Microsoft Word and Excel. Microsoft Excel is for power managing tabular data, but doing them well in both is an essential Microsoft Office skill. Use them at every opportunity.
If you could get paid to play video games, would you do it? It’s definitely possible—thousands of people do it every year—but your options are pretty limited and it may not be as fun as you think. Still interested? Keep reading.
Before we begin, let’s make a distinction between indirect and direct sources of income from video games. Indirect income doesn’t come from a particular game but outside activities that can be applied to any game, such as streaming or video game testing. Direct income comes directly from a particular game.
This article is about direct income. Here are some games that have elements of real money transaction built right into the game’s core design, allowing you to walk away from a gaming session with a few more dollars in your pocket.
CS:GO, Dota 2, Team Fortress 2
Team Fortress 2 (TF2) was Valve’s chosen franchise to spearhead their “cosmetic item market” approach to revenue, who then took it to new levels in Dota 2 and Counter-Strike: Global Offensive (CS:GO). One could say that Valve’s execution of the cosmetic item market idea really drove the success of these games.
In short: cosmetic items (called “skins”) make you look different in-game. Some items look cooler than others, and some items are so cool that players are willing to pay money to acquire them. By engaging in item trading (also known as “skin trading”), you can exchange the items you have (earned by playing the game itself) with another player for real dollars. This is all permitted—even encouraged—by Valve.
This is done through the Steam Community Market, which is an official marketplace run by Valve where you can put cosmetic items up for sale and earn Steam Wallet funds.
CS:GO, Dota 2, and Team Fortress 2 are the most popular Valve games, but the Market supports dozens of other games: PUBG, Rust, Black Squad, Unturned, and more.
Note: Steam Wallet funds cannot be withdrawn, so whatever money you earn can only be spent on Steam games or other items for sale on the Steam Community Market.
If you do want to trade skins for real-life cash, you can use a third-party site like BitSkins (withdrawal by PayPal, Bitcoin, Ethereum, Bank Wires) or StoneFire (withdrawal by PayPal and Bank Wires). These sites take a percentage cut of every transaction. Third-party sites can be risky, so exercise caution when trading!
Skin markets like these are one of the easiest ways to earn money by playing games.
Second Life
Second Life is a uniquely social MMORPG that debuted way back in 2003. Does anyone even talk about it anymore? Hasn’t it fallen off the radar? To my surprise, not only is Second Life still going strong, but it has a strong player-centric economy from which you can make real money.
It all revolves around the in-game currency, Linden Dollars (L$). Players can earn L$ by engaging in various activities in the game. Some activities are passive, such as sitting on camping chairs, while others are active and require certain skills, such as creating virtual content and trading virtual real estate.
Some players would rather save time and spend cash for L$, so you can sell your surplus L$ to other players on the LindeX Currency Exchange for real money. The exchange rate changes based on supply and demand, but tends to hover around L$ 250 for every $1 USD. With time and perseverance, it’s possible to make a profit on the Exchange.
Entropia Universe
There’s another MMORPG in the same vein as Second Life that’s built on a real cash economy: Entropia Universe, formerly known as Project Entropia. Also launched back in 2003, this game isn’t a pure social world like Second Life—there are actual gameplay elements (e.g. hunting, missions, crafting, etc.) so it does feel more like a traditional MMORPG than Second Life.
The in-game currency is called Project Entropia Dollars (PED) and exists at a fixed exchange rate of 10 PED per $1 USD. What’s unique about Entropia Universe is that PED can be directly withdrawn into a bank account, no middleman services necessary.
In a very real way, Entropia Universe is the closest thing you’ll find to a game whose economy is inherently connected to the value of real-world money. But whether the game is fun enough to warrant the effort? Only you can decide.
Other Ways to Get Paid to Play Games
It must be said: regardless of which of these games you decide to play, there’s no guarantee that you’ll actually make a profit on your time—and an even smaller chance that you’ll earn more than minimum wage.
Want to make money playing games? You’ll have to sink a lot of hours in, which can easily turn fun gameplay into hard work. You might hear the occasional story of a player who earns thousands of dollars per month, but they are the exception.
Are there any other games out there that allow players to make money from gameplay? Note that there are several other ways to make money playing games, including streaming, Let’s Plays, podcasts, esports, and more.
Amazon’s Alexa may be in ten thousand different devices now, but they all have one other thing in common: they’re new. So for those of us that prefer old things but still want to be able to set timers and do metric-imperial conversions without pulling out our phones, Grain Design is retrofitting these fabulous old telephones to provide Alexa access with no other hints of modernity. There’s even a privacy angle!
The phones themselves (spotted by a BoingBoing tipster) are genuine antiques, and not even the mass-produced Bell sets you see so often. I personally love the copper-plated model, though I certainly wouldn’t say no to the candlestick.
Dick Whitney, who runs the company, modifies the hardware to make room for an Echo Dot inside. Pick up the phone and speak, and Alexa answers, just like the operators of yore! Except you can ask Alexa anything and it won’t be irritated. Some of the Alexaphones, as he calls them, will include the original audio hardware so you can experience the cognitive dissonance of talking to a virtual assistant and having them answer using a century-old speaker. (I bet it sounds terrible and brilliant.)
“The Echo microphones have their connections severed or are removed completely, and the microphone in the handset is connected via the original switches in the base, so it’s only in contact when the handset is picked up,” explained Whitney in an email.
The modifications to the phones don’t end there: in the rear of each will be a 1/8″ audio port so you can plug in a real speaker. No one wants to sit at their telephone table (remember those?) and listen to a few songs in mono through vintage hardware. Although having written that sentence I do have to say I’d try it once. Right now all the audio would have to go out that way, but Whitney says he may have a trick to switch it back and forth in the future (you can always just unplug the audio for privacy).
There’s also an LED hidden on the front so you have that basic feedback of whether the device is on, listening, and such. The rotary dial isn’t used, unfortunately, though more because it’s hard to apply its principles to a voice-operated device.
“It’s funny,” he wrote when I asked about the latter, “I’d actually built an installation for Android at MWC [Mobile World Congress] a few years ago that used a rotary dialer, so I know how to do it and have the hardware around (it’s very simple), but both couldn’t figure out a function that seemed interesting enough (dial 1 to increase the volume? Certainly open to suggestions) and didn’t want to add more complexity inside the telephones. Maybe in the future!”
No soldering or weird old tech stuff required on your part — the device will run on USB power and set up just like any other Alexa gadget. Of course, these things also cost $1,500. Yeah, kind of out of my price range, too. Still, they’re lovely and a great subversion of the “smart home” idea.
In the immediate aftermath of the walkout, which saw thousands of Googlers across the globe protest the company’s mishandling of sexual harassment and misconduct claims, the search giant said it would put an end to its policy of forced arbitration for employees claiming workplace harassment. Facebook followed suit, announcing the next day that it would allow its employees to pursue claims of sexual harassment in court.
Today, Airbnb and eBay confirmed to TechCrunch they too would no longer require sexual harassment claims to be settled through private arbitration. Their announcements emerged as a result of a BuzzFeed News article exploring which tech companies were updating their policies in light of the Google protest.
Thousands of Google employees protested the company’s handling of sexual harassment & misconduct allegations on Nov. 1.
“We are a company who believes that in the 21st Century it is important to continually consider and reconsider the best ways to support our employees and strengthen our workplace,” a spokesperson for Airbnb said in a statement provided to TechCrunch. “From the beginning, we have sought to build a culture of integrity and respect, and today’s changes are just one more step to drive belonging and integrity in our workplace.”
Here’s what eBay had to say about its decision: “eBay takes great pride in fostering an inclusive culture that allows employees to feel comfortable and encouraged to report any workplace issues. We’ve adjusted our existing employee policy regarding sexual harassment claims to better reflect and encourage eBay’s values of being open, honest and direct.”
According to BuzzFeed, Pinterest, Oath, Twitter, Reddit and others have never required mandatory arbitration. Uber, Lyft and Microsoft each put an end to forced arbitration in the last year.
Arbitration is a private method of solving a dispute without a judge, jury or right to an appeal. Companies that require forced arbitration waive their employees right to sue and to participate in a class action lawsuit.
Throwing out its policy of forced arbitration was one of the five demands disgruntled employees had for Google. In a Medium post last week, the walkout organizers commended Google’s decision to end forced arbitration while emphasizing the company’s failure to address all of their demands.
…”The response ignored several of the core demands — like elevating the diversity officer and employee representation on the board — and troublingly erased those focused on racism, discrimination, and the structural inequity built into the modern day Jim Crow class system that separates ‘full time’ employees from contract workers. Contract workers make up more than half of Google’s workforce, and perform essential roles across the company, but receive few of the benefits associated with tech company employment. They are also largely people of color, immigrants, and people from working class backgrounds.”
Google employees began crafting a plan for a company-wide walkout in late October, days after a bombshell New York Times investigation revealed Google had given Android co-creator Andy Rubin a $90 million exit package despite multiple relationships with other Google staffers and accusations of sexual misconduct.
Rubin, for his part, has said that the NYT’s story contained “numerous inaccuracies.”
Posted by Chong Wang, Research Scientist, Google AI
Speaker diarization, the process of partitioning an audio stream with multiple people into homogeneous segments associated with each individual, is an important part of speech recognition systems. By solving the problem of “who spoke when”, speaker diarization has applications in many important scenarios, such as understanding medical conversations, video captioning and more. However, training these systems with supervised learning methods is challenging — unlike standard supervised classification tasks, a robust diarization model requires the ability to associate new individuals with distinct speech segments that weren't involved in training. Importantly, this limits the quality of both online and offline diarization systems. Online systems usually suffer more, since they require diarization results in real time.
Online speaker diarization on streaming audio input. Different colors in the bottom axis indicate different speakers.
In “Fully Supervised Speaker Diarization”, we describe a new model that seeks to make use of supervised speaker labels in a more effective manner. Here “fully” implies that all components in the speaker diarization system, including the estimation of the number of speakers, are trained in supervised ways, so that they can benefit from increasing the amount of labeled data available. On the NIST SRE 2000 CALLHOME benchmark, our diarization error rate (DER) is as low as 7.6%, compared to 8.8% DER from our previous clustering-based method, and 9.9% from deep neural network embedding methods. Moreover, our method achieves this lower error rate based on online decoding, making it specifically suitable for real-time applications. As such we are open sourcing the core algorithms in our paper to accelerate more research along this direction.
Clustering versus Interleaved-state RNN
Modern speaker diarization systems are usually based on clustering algorithms such as k-means or spectral clustering. Since these clustering methods are unsupervised, they could not make good use of the supervised speaker labels available in data. Moreover, online clustering algorithms usually have worse quality in real-time diarization applications with streaming audio inputs. The key difference between our model and common clustering algorithms is that in our method, all speakers’ embeddings are modeled by a parameter-sharing recurrent neural network (RNN), and we distinguish different speakers using different RNN states, interleaved in the time domain.
To understand how this works, consider the example below in which there are four possible speakers: blue, yellow, pink and green (this is arbitrary, and in fact there may be more — our model uses the Chinese restaurant process to accommodate the unknown number of speakers). Each speaker starts with its own RNN instance (with a common initial state shared among all speakers) and keeps updating the RNN state given the new embeddings from this speaker. In the example below, the blue speaker keeps updating its RNN state until a different speaker, yellow, comes in. If blue speaks again later, it resumes updating its RNN state. (This is just one of the possibilities for speech segment y7 in the figure below. If new speaker green enters, it will start with a new RNN instance.)
The generative process of our model. Colors indicate labels for speaker segments.
Representing speakers as RNN states enables us to learn the high-level knowledge shared across different speakers and utterances using RNN parameters, and this promises the usefulness of more labeled data. In contrast, common clustering algorithms almost always work with each single utterance independently, making it difficult to benefit from a large amount of labeled data.
The upshot of all this is that given time-stamped speaker labels (i.e. we know who spoke when), we can train the model with standard stochastic gradient descent algorithms. A trained model can be used for speaker diarization on new utterances from unheard speakers. Furthermore, the use of online decoding makes it more suitable for latency-sensitive applications.
Future Work
Although we've already achieved impressive diarization performance with this system, there are still many exciting directions we are currently exploring. First, we are refining our model so it can easily integrate contextual information to perform offline decoding. This will likely further reduce the DER, which is more useful for latency-insensitive applications. Second, we would like to model acoustic features directly instead of using d-vectors. In this way, the entire speaker diarization system can be trained in an end-to-end way.
To learn more about this work, please see our paper. To download the core algorithm of this system, please visit the Github page.
Acknowledgments This work was done as a close collaboration between Google AI and Speech & Assistant teams. Contributors include Aonan Zhang (intern), Quan Wang, Zhengyao Zhu and Chong Wang.
When you want to move or order a row around, use the combination of Alt+Shift+Up Arrow and Alt+Shift+Down Arrow to order the row up or down. Move contiguous rows by selecting them all first.
