EU PASSES RULES RESTRICTING USES OF ARTIFICIAL INTELLIGENCE

On 11 May, the European Union’s (EU’s) parliament passed a sweeping set of controls that will be incorporated into the union’s Artificial Intelligence Act, likely to become the world’s first comprehensive body of regulations governing AI.

The rules require that AI developers report content their AIs create and disclose the sources of materials used to train an AI so that the creators of the source materials can be compensated.

The rules also ban the use of AI as a tool in facial recognition to surveil EU citizens.

The passage ends two years of squabbling among the parliament’s members over the kinds and scope of rules to be enacted.

Next, the rules will be debated among EU member states and reviewed by the European Commission.

The debate will include consideration of using AI for facial recognition under special circumstances, such as terrorist threats, persons familiar told the Financial Times.

Discussions will be “tough,” in part because “governments are speaking clearly about their willingness to use these systems,” parliament member Brando Benifei said to the FT.

The measure was the most debated in EU history and sorted through more than 1,000 proposed amendments, according to Dragos Tudorache, who led the effort.

Concerns over the sudden rise in AI’s power and possible applications also have stirred in the U.S.

Last week, the U.S. Federal Trade Commission announced it is “focusing intensely on how companies may choose to use new AI technology, including generative AI tools, in ways that have actual and substantial effects on consumers.”

TREND FORECAST: As has been the case with new technologies since the dawn of radio, politicians and regulators will argue about controls while the technology spreads, morphs, and alters social norms and behavior.

While governments will try to regulate AI, and over time a patchwork of rules will accumulate, that in the end, as we had forecast in one of our Top Trends 2023: AI WE OWN YOU.

ARTIFICIAL INTELLIGENCE UPENDS WALL STREET

The share price of chip maker Nvidia has shot from $191 in February to $283 last week and now trades at 164 times its past 12-month earnings. Chegg, an educational publisher, watched its stock price fall from $17 to below $10 during the last week of April.

Microsoft’s market cap has exploded by another $500 billion since January when it invested $10 billion in OpenAI, the industry’s leader. Alphabet, Google’s parent, has added 22 percent to its share value this year on the strength of Bard, its version of AI.

This is the U.S. stock market under the spell of artificial intelligence (AI). Everyone knows it will change everything but no one, including the companies whose shares are being traded, knows exactly how yet.

The markets’ attention has been riveted on AI by ChatGPT, a form of generative AI created by OpenAI, the firm backed by Microsoft and Elon Musk, among other tech luminaries.

A Chatbot precursor was released in June 2020, but when Chat GPT was unveiled last 30 November, it booked 100 million users in its first two months, making it the fastest-growing app on record, according to Goldman Sachs.

However, there is broad agreement that AI will reflect the far-ranging ripples that past technologies have made.

Online streaming services killed the video rental industry and Blockbuster; cameras in smartphones did away with Polaroid and photo developing services and drove Apple’s ascendance.

AI’s “longer-term implications are greater than we can imagine,” chief strategist Michael Green at Simplify Asset Management told The Wall Street Journal.

“You just don’t know all the knock-on effects,” Will Graves, Boardman Bay Capital Management’s chief investment officer, said to the WSJ. “Nobody saw that Uber was coming out of the iPhone to hammer the taxi industry”—and, as it turns out, to discourage an entire generation from owning cars. 

TRENDPOST: There were more than 300 references to “generative AI” in corporate conference calls around the world so far this year, data service AlphaSense said, while being virtually neglected in calls before January.

Generative AI can answer questions and create videos, natural language text, and other media content. 

So far, ChatGPT has invented new drugs and materials, delivered answers to knotty questions in biochemistry, and begun replacing entry- and low-level workers in everything from fast-food restaurants to accounting firms.

Its impact has been so broad and so instantaneous that investors are still trying to comprehend what its practical impact will be across the range of industries.

GOOGLE REVAMPS SEARCH ENGINE TO INCLUDE AI 

At its annual conference for developers last week, Google announced a range of new products and upgrades related to Bard, its ChatGPT-like version of generative artificial intelligence (AI).

AI is becoming part of Google’s search engine, Android mobile operating system, and cloud computing service. Some of the features are in place now, while others will spread slowly through beta platforms Google is calling “Search Labs.”

The features are based on Google’s PaLM-2 [stet], its new AI large language model that the company has embedded in its email software and that developers can use to train their own AIs.

The search engine will offer brief texts of answers to queries as well as links to other Internet pages. The feature also will allow users to converse with the search engine in natural spoken language.

Google’s AI responses will include sources of the information so users can judge its validity—and also will include ads, Cathy Edwards, vice president of Google Search, told the Financial Times.

Duet AI is Google’s new tool that helps create emails and spreadsheets, similar to Microsoft’s “Copilot” feature incorporated into its Office suite.

Google also has partnered with German software giant SAP SE to engineer a way for AI to analyze data.

Google is chasing Microsoft, which in February premiered a version of its Bing search engine incorporating ChatGPT, which Microsoft helped develop. In March, the company OpenAI—which created ChatGPT—unveiled its new GPT-4 AI, which also can be used through Bing.

The moves were seen as Microsoft’s attempt to raise Bing to dominance among search engines. Internet searches account for most of Google’s revenue.

Microsoft’s February bombshell erased billions from Google parent Alphabet’s stock value, dropping it 12 percent within days. 

In response, Alphabet combined its Deep Mind and Google Brain AI R&D programs to speed its new feature to market.

The conference “quelled near-term investor angst over [Alphabet’s] lack of AI innovation,” Jeffries analyst Brent Thill wrote in a note to clients. “Google presented a much more coherent message about its generative AI strategy.”

TRAINING AI TO SAY, “YOU WANT FRIES WITH THAT?”

In June, burger chain Wendy’s will begin testing an AI chatbot that will take drive-through orders and answer simple, frequently-asked questions from customers.

The chatbot will run on Google Cloud’s AI software and will be as easy and natural as talking with a human, the company has promised.

The chatbot will “deliver a truly differentiated, faster, and frictionless experience for our customers,” CEO Todd Penegor said in a statement.

Drive-through ordering surged in popularity during the COVID era. About 80 percent of its customers prefer ordering that way, Wendy’s said.

The company is the latest to adopt AI to undertake simple, routine tasks as the fast-food industry continues to grapple with labor shortages and the need to pay higher wages.

“I don’t think in three years, there’s going to be a drive-through having a human take your orders,” Krishna Gupta, chairman of Presto Automation, which provides an AI ordering system for restaurants, said in a Bloomberg Television interview earlier this month.

However, “we do not anticipate reducing labor, but instead shifting crew responsibilities to meet the increased volume of Wendy’s orders expected in the drive-through and across our growing digital channels,” the burger company said.

TREND FORECAST: AI will reduce the number of jobs for teens and low-skill workers not only in fast food but in every industry.

Every employer of any size already is scheming ways to use AI to cut costs, which for many, and perhaps most, companies means cutting workers off payrolls. That can mean accountants and attorneys may be no safer than the clerks who make out their paychecks.

AI will ripple through the world’s employment structure at freeway speeds through the rest of this year and for years to come.
The Trends Journal will report on and forecast developments in this technology that will be even more transformative than the personal computer revolution.