Tech startup makes telemarketers sound more American
Like AI, facial recognition, and robotics, voice technology is one of the hottest commodities in tech. As intelligent smart speakers get people more accustomed to talking with machines, the revenue potential of voice tech rises. But as we’ve seen before, the tech industry doesn’t always prioritize the social implications of new systems.
Silicon Valley-based startup, Sanas, has an interesting yet questionable mission: to make call center workers sound more American no matter where they are from. Their website states, “Our mission is to make lives better by expanding the horizons of what is possible with your voice.” However, how they are pursuing this mention is considerably cringeworthy.
In a demo on Sanas’ homepage, a fellow with an Indian accent reads a standard call center script with a recognizable accent for English speakers from India. When you click the slider, his voice transforms to a highly Americanized pronunciation that, as some have pointed out, also sounds more “white.”
“We don’t want to say that accents are a problem because you have one,” Sanas president Marty Sarim said in an interview. “They’re only a problem because they cause bias and they cause misunderstandings.”
The company depicts its approach as “accent matching,” and claims it can “improve understanding by 31% and customer satisfaction by 21%.” The technology offers multiple accent changes, according to Sanas, but the demo only includes an Indian accent morphing into typically American.
One of the biggest criticisms of Sanas’ new technology is that digitally altering a person’s natural accent does nothing to reduce prejudice and systemic bias in America toward non-native speakers. Others have pointed out that, while the developers claim their tool is “a step towards empowering individuals, advancing equality, and deepening empathy,” call center employees are often micromanaged, with little autonomy. Critics maintain that digital accent-washing does nothing to improve stressful call center work conditions or eliminate hostility from angry consumers.
In other news
- Hackers stole LastPass source code. Malicious actors hacked into the LastPass developer systems and came away with proprietary technical information in addition to portions of its source code. According to ZDNET, the hackers gained access to the system through a single compromised developer account. In a blog post, LastPass CEO Karim Toubba wrote that there is no evidence that the hackers accessed customers’ personal data or their encrypted password vaults. The company claims it has seen no evidence of further malicious activity and has since enlisted the help of a leading cybersecurity and forensics firm to find out exactly what happened.
- Chemists find a method to destroy harmful chemicals present in everyday objects. The BBC reports that there is a new, low-cost way to permanently destroy poly- and perfluoroalkyl substances (PFAS), which have been historically difficult to eradicate. PFAS are fluorine-based compounds used to create various household objects, such as non-stick cookware, make-up, and adhesives. Unfortunately, they are linked to health risks like cancer and birth defects at certain levels. Research has shown that low levels of PFAS have been found in rainwater worldwide, and scientists are concerned that increased levels could contaminate the soil, drinking water supplies, and wildlife. While it is recommended that PFAS are only used in products and processes that are vital for society and where there are no alternatives, PFAS are still widely used worldwide.
- You can now speak with Google’s LaMDA chatbot. It’s been a couple of months since Google engineer Blake Lemoine claimed that the tech giant’s Language Model for Dialogue Applications (LaMDA) is sentient, sparking a great deal of debate. In the end, Google accused Lemoine of anthropomorphizing the bot. Now, Google is allowing the general public to see for themselves. According to Gizmodo, Google will open its AI Test Kitchen app to the public. However, access to the chatbot won’t be entirely open-ended. Instead, the bot will be presented through sets of structured scenarios. Anyone interested can register here.
- How to orchestrate the biggest Rickroll ever. It’s been 15 years since the first ever Rickroll, and its popularity shows no signs of letting up, likely making it the Internet’s most enduring meme. Case in point? Four high school students succeeded in hijacking 500 screens in six school buildings to play Rick Astley’s “Never Gonna Give You Up” as a graduation prank. According to Wired, the prank took place on April 20, 2021, and required the pranksters to carry out a series of illegal hacking techniques to pull off the prank. These included breaking into the school’s IT systems, writing and executing custom codes and scripts, and secretly testing them on the school’s IT system at night.
- High-powered X-ray may help us understand early printed documents. A team of scientists at the SLAC National Accelerator Laboratory in California recently examined two of the earliest printed documents with a powerful X-ray. The aim is to find out what kind of metal types printed the documents and the chemistry of their ink. Gizmodo reports that one of the documents is two pages of a Gutenberg Bible, the first major book to be mass printed using one of the earliest moveable-type printing presses in Europe. The second document hails from Korea, consisting of pages from the 580-year-old Spring and Autumn Annals. It was also printed via a moveable-type printing press and preceded Gutenberg’s work by several years.
Tip of the week: Can you trust password managers?
This week hackers accessed source code from password management software LastPass. This kind of thing might make you question whether or not it’s a good idea to trust a company — any company, including LastPass, 1Password, or even Google and its own Password Manager — with the keys to the kingdom After all, if a bad actor gains access to your account login details, they could take over your website or social media accounts, drain your bank account, send out emails pretending to be you, and other unspeakable horrors.
So can you trust password managers? Here are some things to keep in mind.
- Choose a reputable company. The best companies will keep your passwords and other data (including your master password) encrypted, so even in the unfortunate case of a security breach, hackers will have nothing more than a bunch of gibberish.
- Employ strong master passwords, especially on Google. If someone can guess your password and log in as you, all the encryption in the world won’t save you.
- Be smart about where you store master passwords. Use a passcode on all electronic devices (including your work computer, in compliance with your IT department) so no one can access your information without authorization. And don’t store passwords in a notebook or under the keyboard — which you should already know if you use a password manager.
- Be careful about free password managers. While many are potentially as secure as their paid counterparts, they may have limits on data or levels of security, only work on a single device, or have other limitations. And keep in mind: you get what you pay for.
In the end, no password management system is foolproof. But in the end, we believe that the top companies do everything in their power to keep your data secure, while the alternatives, such as spreadsheets or sticky notes, are never as secure or convenient.