Add this to the AI-for-public-good category: A mobile-phone company in the UK has created a digital AI granny called Daisy to thwart phone spammers by keeping them on the line for a long time to prevent them from calling more people.
The talkative chatbot, created by O2 and modeled on the grandmothers of some of the bot’s developers, is the newest member of O2’s fraud prevention team, the phone company says in a post about Daisy. You can also watch a minute-and-a-half-long video of Daisy in action, in which she says, “Hello, scammers. I’m your worst nightmare.”
O2 describes her as “completely indistinguishable from a real person” and says Daisy can “interact with scammers in real time without any input from her creators.” She’s kept scammers on calls successfully for 40 minutes at a time by pretending not to be tech savvy, talking about her cat Fluffy and generally wasting “as much of their time as possible with human-like rambling.”
Though in the past year AI helped the US government recover $1 billion lost to check fraud scams, consumers around the world lost more than $1 trillion to online scams, according to the Global Anti-Scam Alliance, CNN reported.
Here are the other doings in AI worth your attention.
AI is stressing out Gen Z
Another week, another dozen or so reports about what people think of AI and how companies are incorporating generative AI into the workplace. I’ll call out just a few.
First, Gallup found that nearly two-thirds of chief human resource officers at Fortune 500 companies say that AI will start “replacing roles in their organization within the next three years.” At the same time, the research company noted a problem with upskilling, or how much companies are/aren’t investing in AI training programs. Gallup said most employees feel “they are not skilled enough to do their jobs well today and are not getting encouragement they need to be ready for the future.” That may be because only about a quarter of US employees feel their companies encourage them to learn new skills, according to Gallup surveys.
Intuit took a more upbeat look at the future of jobs and called out the AI roles that will be most in demand. Unsurprisingly, eight are engineering related: AI engineer, computer vision engineer, machine learning engineer, data engineer, deep learning engineer, robotics engineer, software engineer and natural language processing (NLP) engineer. But companies will also need data scientists and business intelligence experts to create and manage dashboards, reports and data visualization tools, Intuit added. In addition, employers will be looking for product managers who can oversee the rollout of AI tools and services, and AI ethics specialists, people who can help develop ethical frameworks and guidelines.
Meanwhile, Atlassian surveyed 5,000 “knowledge workers” in the US, Australia, France, Germany and India and identified five types of AI “mindsets.” Nine percent of those surveyed said they feel AI is “useless” in the workplace. People characterized as Stage 1 thinkers, about 29% of those surveyed, described AI as a “tool they can use sometimes, but just to accomplish a specific task.” Stage 2 thinkers, 30%, consider AI a personal assistant able to help them do their job. Stage 3 thinkers, or 21% of workers polled, see AI as a “creative partner,” while 12% fell into Stage 4, describing AI as akin to a “team of experts who can enhance their decision making.” Most workers today are in Stage 2 and 3, Atlassian said.
And last but not least, EduBirdie, a professional writing service, asked its Gen Z users what they think of AI, as part of an ongoing set of surveys. After concerns about a global war, the climate, a major economic collapse and another pandemic, Gen Zers named AI as among their top worries, according to a survey of 2,000 who were asked to describe what’s stressing them out. Ten percent said they’re worried about AI “taking over and harming humanity.” The findings are here. No word on how many Gen Zers have seen The Terminator.
Kissinger says AI can aid humanity — with government guardrails
Henry Kissinger is dead, to begin with (yes I borrowed that line from Dickens).
But that hasn’t stopped some of his final thoughts from being shared, using a re-creation of the diplomat’s voice via ElevenLabs AI voice simulator. And that’s actually fitting, because the tech is giving voice to Kissinger’s thoughts on AI, as he wrote them in his final book, Genesis: Artificial Intelligence, Hope, and the Human Spirit. Co-written with former Google CEO Eric Schmidt and Microsoft Chief Research Officer Craig Mundie, the book explores AI and its “profound implications for humanity,” and you can “hear” Kissinger, and his very much alive co-authors, in this 12-minute video promo for Genesis.
With AI, people are wielding a power we can’t possibly understand, warns Kissinger, who served as secretary of state for President Richard Nixon. Still, he sees its potential — as long as there are proper guardrails. “It is imperative that governments create an environment where ethical considerations and technological advancement can progress in tandem,” according to Kissinger. “If we create a proper framework, artificial intelligence holds remarkable promise for advancing human progress and addressing some of the most pressing challenges of our time.”
Mundie and Schmidt also share their optimism and concern, with the important reminder that we — people — are making AI. Notes Schmidt, “The technology we’re inventing is valueless — it doesn’t have human values in it. It’s just technology. It could be used for good or evil. It’s trained by human systems, by human language, by human construction.”
Also worth knowing…
We’ll see how the US approaches AI, now that a new administration will take over in January. In a scoop, Axios reported last week that the Trump administration is “considering naming an AI czar in the White House to coordinate federal policy and governmental use of the emerging technology.” The czar won’t be Trump advisor Elon Musk, who may have a say in the administration’s pick, Axios added.
In a small study of 50 doctors, researchers found that GhatGPT didn’t really improve doctors’ ability to diagnose their patients’ illnesses compared with “doctors who used only traditional resources,” The Washington Post reported. But the study also found that “ChatGPT on its own performed better than either group of physicians,” the Post noted.
