AI & Automation News Roundup from Uniphore - October 21

From exciting, new applications to electrifying tech innovations, the twin fields of artificial intelligence and automation are evolving daily. If you missed the latest developments making the news, we have you covered. Here’s a quick recap of the top three AI and automation stories worth knowing.

White House Unveils New “AI Bill of Rights”

Earlier this month, the White House released its Blueprint for an AI Bill of Rights, which outlined five core principles for designing, using and deploying automated systems ethically. The announcement fulfills a promise made last year by the Office of Science and Technology Policy (OSTP) to draft a comprehensive list of protections against malicious, discriminatory and ineffective AI. Critics, however, were quick to note that the voluntary guidelines are largely toothless without the rule of law to enforce them.

The five principles outlined in the document are:

1. Protection from unsafe or ineffective systems designed with an intent or reasonably foreseeable possibility of endangering the safety of users or their communities.
2. Protection from algorithmic discrimination based on race, color, ethnicity, sex, religion, age, national origin, disability, veteran status, genetic information, or any other classification protected by law.
3. Protection from abusive data practices, including enhanced restrictions for sensitive domains (i.e. health, finance, criminal justice) and protection from unchecked surveillance.
4. Plain-language transparency, via notices and explanations, regarding how automated systems are being used and how and why they impact users.
5. The ability to opt-out of automated systems in favor of a human alternative, where appropriate.

“This document is laying down a marker for the protections that everyone in America should be entitled to,”  Alondra Nelson , OSTP Deputy Director for Science and Society, told The Washington Post. “This is a particularly important example of how our concerns and hopes for automated technologies really meet some of the other highest priorities of this administration, like workers’ rights.”

Anand Mahindra: Humans Only Make Up 38.5% of Internet Traffic

In a tweet that has since gone viral, industrialist and Tech Mahindra Chairman, Anand Mahindra, stated that, "Humans only make up 38.5% of internet traffic. The other 61.5% is non-human (bots, hacking tools, crypto, etc)". Mahindra’s October 4 post reiterated an earlier statement made by World of Statistics that elicited a similar reaction, including a shocked response from Tesla CEO Elon Musk.

While certainly bold, the number reported by World of Statistics (and shared by Mahindra) isn’t unique. A security report from Barracuda published earlier this year drew a similar conclusion, stating that humans represent only 36% of internet traffic. Barracuda broke down the remaining 64% into automated traffic from bad bots (39%) and good bots (25%), such as search engine crawlers and social media bots. Of the bots that make up all automated internet traffic, “60% are dedicated to carrying out malicious activities at scale,” says Nitzan Miron, VP of Product Management, Application Security for Barracuda. “When left unchecked, these bad bots can steal data, affect site performance, and even lead to a breach.”

NOTE: Tech Mahindra is a featured Uniphore partner. You can learn more about our joint commitment to drive innovation in customer experience here. You can also view our webinar, “Reimagine the IT Help Desk with Conversation AI”, featuring Lawrence Sathiaraj, Global Delivery Head at Tech Mahindra, here.

AI Could Soon Diagnose Illnesses from Vocal Data

The human voice is a goldmine of data. We already know that modern conversational AI can accurately identify a speaker’s intent, sentiment and even emotional state based on their vocal patterns and inflections. Now, scientists hope to use that same vocal data to diagnose a range of illnesses—from neurological disorders to certain types of cancer.

The National Institute of Health recently announced its plans to build a massive database of human voices for use in AI-based biomedical and behavioral research. As part of its $130 million Bridge2AI program, the NIH will take ethically sourced vocal samples from scores of donors, then train AI to analyze the dataset for key biomarkers, looking at everything from vocal cord vibrations to breathing patterns. For example, someone with an irregular breathing pattern may show early signs of Parkinson’s Disease. (We discussed similar research in last month’s news roundup.)

Speaking with NPR, Dr. Yael Bensoussan , director of the University of South Florida's Health Voice Center and a leader on the study, explained the project’s goal: “We asked experts: If you close your eyes when a patient comes in, just by listening to their voice, can you have an idea of the diagnosis they have?"

Bensoussan’s team hopes to use the vocal database to ultimately power an app that could help serve traditionally underserved communities. Long-term, the data could potentially interface with vocal assistants like Alexa or Siri to monitor users’ speech for signs of illness and recommend medical attention, if needed.

Ask Your Conversational AI & Automation Questions:

We are bringing the community together to showcase our latest AI & automation innovations and answer your questions. RSVP to get notified when we go live.

From the Experts:

In a recent interview on Conversations that Matter, a podcast by Uniphore , we asked CX and technology expert Kevin Zyskowski what executives look for when approving technology like AI. Here is what he said based on senior positions at JP Morgan Chase, Progressive Insurance and more:

Follow Conversations that Matter, a podcast by Uniphore for more conversations with experts.

Like this newsletter? Comment below and share it with your community.

Did we miss a news story? Email social@uniphore.com or comment below.