Businesses actively use artificial intelligence, but consumers do not trust it

Since late last year, when OpenAI introduced its next-generation chatbot ChatGPT, the number of online queries about artificial intelligence (AI) has grown by more than 1,000%. Many large companies are already actively using new AI technologies in their work and bringing them to the mass market. However, researchers note that consumers are very wary of new AI services. Even those who were optimistic about these innovations last year have begun to change their attitude towards them.

1000% growth

Last week, research company Indusface presented results of a survey on attitudes towards artificial intelligence in British business. Researchers note that since November last year, when OpenAI introduced the ChatGPT chatbot, Internet search queries on the topic of AI have skyrocketed by 1,700%, and the number of chatbot users has exceeded 100 million.

To find out the attitude of business to the new technology, Indusface surveyed about 2 thousand people in the UK in September, working in companies from 25 industries – from advertising to industry.

• Most often, a chatbot is used to prepare reports—27% of respondents reported this.
• In second place is assistance in translating texts from/to foreign languages ​​- 25% of respondents.
• Another 17% said they use AI for research purposes, and 11% even use it to write emails to clients.

The survey revealed that ChatGPT is used more actively in the advertising industry – 39% of respondents said that they use it in their work. In second place are legal services – 38% (see table).

The popularity of AI in the legal field has already led to one of the largest M&A deals of this year – in June, Canadian media company Thomson Reuters announced on the purchase for $650 million of the Californian legal startup Casetext, which has developed a smart assistant using AI. The CoCounsel assistant, launched this year, runs on the GPT-4 platform. It analyzes legal documents, contracts and litigation materials in minutes. Currently, the clients of Casetext, which employs only about 100 people, are more than 10 thousand law firms and legal departments of different companies.

Entrepreneurs undertake

Many large companies have announced the integration of AI into their business and services this year. Travel service Expedia for several months now uses ChatGPT. Users can find out the best routes and travel times for them in a casual conversation by asking questions like: “When is the best time to go to Hawaii for a honeymoon?” and “Is it possible to kite in Egypt in November?” The chatbot instantly gives a detailed, coherent answer, as if a travel agency employee was in front of the client – explaining exactly what the weather is like in a particular region at a particular time, for what type of vacation a particular area is suitable, etc.

Application for learning foreign languages Duolingo Also uses GPT-4, which in its free version gives tasks for translation, reading, pronunciation and listening, and then checks the student’s answers. In a more advanced paid version Duolingo Max The student has access to two new features and exercises: Explain My Answer and Role Play. Explain My Answer challenges students to learn more about their answer in class (whether the answer was correct or incorrect). By clicking a button after certain exercise types, students can enter a chat with Duo to receive a simple explanation of why their answer was correct or incorrect and ask for examples or further clarification. “Role Play” allows students to practice their communication skills with virtual characters: challenges that can earn points are completed in parallel with the character.

Students can discuss plans for future holidays, make a reservation at a restaurant or cafe in the country of the language they are learning, go shopping in the country of the language they are learning, etc.

Coca-Cola cooperates with one of the largest consulting companies in the world, Bain & Company, in the field of marketing. As part of this project, companies are using ChatGPT and an image generator DALL-E to create personalized advertising texts, advertisements and banners.

Consumers don’t trust

Indusface researchers note that despite the popularity of AI in business, people are wary of this technology. 55% of company employees surveyed said they “don’t trust working with another business using ChatGPT or a similar AI platform in their workplace.”

Other researchers have come to similar conclusions. Last summer, PwC interviewed 53.9 thousand workers from 46 countries who, among other questions, answered the following question: “What impact will AI have on your career in the next five years?” Respondents could choose several answer options.

• The most popular option (31%) was: “AI will help me become more productive.”
• In second place (27%) is the option: “AI will create opportunities for gaining new experience and knowledge.”
• The third most popular option (22%) was: “AI will not have any impact on my work.”
• 18% chose the more alarming option: “AI will require new knowledge/qualifications from me that I am unlikely to be able to master.”
• 14% said “AI will change the nature of my job in a negative way,” and 13% said “AI will replace me in the workplace.”

In mid-September, the American research organization Miter published results of a survey on attitudes towards artificial intelligence organized jointly with the sociological organization Harris Poll. The survey was conducted in July among US adults. This is the second such survey conducted by these organizations—the first took place in November last year. The survey revealed that only 39% of respondents consider today’s AI technologies safe – this is 9% less than in the previous survey. The majority of those surveyed also said they were now more concerned than excited about these technologies.

When it comes to AI regulation, 85% of those surveyed would support efforts by government, business and scientists to make it safer. This is 11% more than last year.

Also, 85% of respondents would like the technology industry to make the use of AI more accessible, understandable and transparent to the general public – and before introducing a specific product to the mass market.

On the technology side, concerns are no less serious – 80% are concerned about AI being used to carry out cyberattacks, 78% are concerned about the theft and misuse of personal data, and 74% are concerned about AI being used to create fake political ads. While last year, 54% agreed that AI was sufficiently advanced to be used in areas such as defense and national security, now significantly fewer respondents share this point of view – 46%.

“People’s wariness about AI is not surprising given the potential impact of such technologies on jobs, as well as news about how AI is sometimes used to create deepfakes, fake videos, and the like,” said Douglas Robbins, vice president of Miter. “As the public becomes more aware of AI-enabled products this year, they are increasingly concerned about the practical impact of such technologies on the areas that matter most to people—jobs, healthcare and medicine, transportation, entertainment, and so on.” “

Evgeny Khvostik