It was only a few months ago in November that OpenAI released their ground-breaking ChatGPT chatbot, and since then the chatbot has enjoyed unprecedented success with more than 100 million monthly active users. Inevitably, with the advances in AI technology and the developments within open source AI models, users soon began to shift from traditional website navigation to both generative AI with real-time information, and their app for iOS systems. Now, the first speed bumps are appearing within the success of ChatGPT, with recent data showing a notable decrease in worldwide desktop and mobile traffic.
According to analytics firm Similarweb, in June there was a 9.7% decrease from the previous month in desktop traffic, and a 5.7% decrease in mobile traffic. The amount of time visitors spent on the website dropped by 8.5%. Senior Insights Manager at Similarweb, David Carr believes this is indicative of the chatbot’s novelty beginning to wear off. These stats were accompanied by a decrease in app downloads, more specifically those of ChatGPT, says RBC Capital Markets analyst Rishi Jaluria. Despite their phenomenal growth, they now appear to be hitting a plateau.
The decrease in traffic has been echoed by Bank of America analysts, who explain that visits to ChatGPT were down around 11% per week in June, to 51 million from the previous month. BofA analyst Justin Post believes that app downloads of both ChatGPT and Microsoft Bing (which includes a ChatGPT-based chatbot in the US) decreased 38%, with Bing’s market share decreasing 40 basis points to around 2.8%. This was juxtaposed by Google which has increased their search engine market share very slightly, now over 92%.
The plateauing of ChatGPT usage has raised questions of investing risk for companies such as Google and Microsoft, whom have invested millions in the development and usage of language models. Justin Post has commented on this, saying “as the technology may not seriously threaten Google’s dominance in search and companies may have to find other applications for LLMs, such as in new advertiser tools”.
Despite this, there is still hope for ChatGPT and the bigger AI models being used across the world. OpenAI’s ChatGPT offers a premium subscription service for $20 a month to access GPT-4, and 1.5 million people have already signed up. The Bank of America analysts also speculate that the usage numbers could be on the up when the app they are planning to launch for Android in July is released.
As OpenAI’s Chief Executive Sam Altman divulges, running a service this popular takes an immense amount of computing power, and they generate ample revenue from numerous avenues such as API access to their AI models and their premium subscription service. It is therefore unsurprising that OpenAI are projecting $200 million in revenue this year, despite the slight dip in desktop and mobile traffic.
Overall, the figures would appear to show a slacking of usage with ChatGPT, yet this was inevitable as soon as machines were able to generate real-time information. As our AI technology continues to blossom and develop, it will be interesting to see how ChatGPT and the larger language models progress, and if the users will soon start to grow again or if the plateau will remain.
PYMNTS discusses the increasing popularity of connected cars and the in-car payment infrastructure that is transforming the driving experience. Mercedes-Benz Financial Services USA recently launched an in-car payment feature that allows drivers to identify and reserve parking spaces at off-street parking facilities using their advanced multimedia infotainment system. This service, powered by Mercedes pay, is available on any Mercedes-Benz vehicle in the U.S. and is expected to be useful during the holiday season when parking restrictions are enforced. PYMNTS also mentions other companies in the automotive industry, such as Car IQ, J.P. Morgan, General Motors, and BMW, that are embracing digital payment infrastructure to cater to the growing number of connected cars on the road.
In addition, PYMNTS highlights a recently announced pilot program by Mercedes and Microsoft. They have integrated the ChatGPT generative AI function with the MBUX Voice Assistant in over 900,000 vehicles. This integration allows for improved natural language understanding and expands the topics to which the voice assistant can respond. The collaboration between Mercedes and Microsoft follows a previous partnership between the automaker and Google, which enables drivers to access information about businesses via Google’s Place Details function. Overall, PYMNTS emphasizes the increasing integration of payment and AI technologies in the automotive industry to enhance the convenience and efficiency of driving experiences.
Unit21 has raised $45 million in a Series C round to support the growth of the Fintech Fraud DAO consortium. The consortium aims to combat fraud and money laundering through collaboration and the sharing of data among FinTechs. Unit21’s risk and compliance infrastructure has seen rapid adoption, with the number of transactions it monitors for clients growing fivefold in 2022. The funding round was led by South Park Commons and Tiger Global.
PYMNTS also mentions other industry players joining forces to fight fraud. Pay.UK is piloting a payments industry coalition to bring together data from retail banks and building societies for fraud detection and prevention. Fraud prevention firm Sardine has launched an independent consortium called SardineX, which allows financial institutions and FinTechs to share data and verify the credentials of fund senders or receivers. Plaid has also announced a collaborative network called Plaid Beacon, enabling the sharing of information about stolen identities and compromised accounts among FinTechs, financial institutions, and Plaid-powered customers.
According to an article on pymnts. com, cryptocurrency miners are finding new opportunities in the artificial intelligence (AI) industry. The report mentions that companies in the crypto business, such as Satoshi Spain, have shifted their focus to AI. Satoshi Spain, which used to sell crypto-mining rigs, has repurposed its machines for AI applications. PYMNTS explains that the graphic chips used in these computers, known as graphics processing units (GPUs), are useful for both cryptocurrency mining and training AI systems. After the ethereum network stopped mining new coins, millions of GPUs became available for the AI industry. PYMNTS also highlights the potential collaboration between AI and crypto, with cryptocurrency platforms making AI systems more accessible and allowing more players to contribute to its development.
In addition, PYMNTS mentions a report by Coinbase, which argues that AI has the potential to reshape the crypto sector. Despite a decline in AI investments, the report suggests that the intersection of AI and crypto presents an important opportunity for entrepreneurs. PYMNTS points out that collaboration between industry and regulators is crucial for the growth of the AI industry and the expansion of digital influence. Overall, PYMNTS highlights the increasing interest and potential for collaboration between AI and cryptocurrency.
IKEA is using AI technology to train call center employees to become interior design advisors. The Swedish furniture company aims to expand its home improvement services while using an AI bot named Billie to handle routine customer inquiries. IKEA has already extended its interior design services to the UK and the US, offering video call consultations and product recommendations for a fee. Since 2021, IKEA franchisee Ingka has trained 8,500 call center workers to serve as interior design advisors, while Billie has effectively managed 47% of customer queries directed to the call centers over the past two years.
The implementation of AI systems in retail organizations is leading to potential reductions in headcount as certain job roles become redundant or automated. AI-powered automation can streamline tasks such as inventory management, data analysis, and customer support. However, IKEA claims that it is not currently seeing an impact on its workforce due to increased AI utilization. The introduction of generative AI is redefining customer service, enabling AI to detect emotions, provide guidance, facilitate transactions, and prevent customers from repeatedly shouting “representative” into their phones.
According to an article on PYMNTS. com, the Federal Trade Commission (FTC) is preparing to file an antitrust suit against Amazon. The suit focuses on Amazon’s online marketplace and alleges that the company uses its power to punish merchants who refuse to use its logistics services while rewarding those who do. The suit centers on third-party sellers, who make up 60% of Amazon’s eCommerce sales, and claims that Amazon penalizes them for not using certain services, such as advertising and shipping. The FTC is also investigating the algorithm that selects merchants for Amazon’s “Buy Box.” PYMNTS notes that FTC Chair Lina Khan is unlikely to agree to a compromise and may push for a court to restructure the company.
Additionally, PYMNTS mentions that the FTC recently sued Amazon over practices related to Amazon Prime recurring subscriptions. The agency claims that Amazon employed “dark patterns” to deceive consumers into enrolling in the program and subjected them to a complicated cancellation process. Furthermore, Amazon is reportedly facing a full-scale EU antitrust investigation regarding its acquisition of iRobot, the maker of Roomba vacuum cleaners. The European Commission has scheduled a four-month probe into the deal.
According to PYMNTS, Google is laying off employees at its mapping app Waze and merging it with its own map products. The company is reducing roles focused on Waze Ads monetization and providing employees with mobility resources and severance options. This move comes as Google is under pressure to reduce costs and merge its mapping teams. The layoffs come after Google parent Alphabet announced job cuts across its product areas. Additionally, Google reported mixed results in advertising, with a 1% decrease in advertising revenue growth during the first quarter of 2023, but a 1.7% growth in revenue for its search advertising division.
In summary, PYMNTS discusses Google’s decision to lay off employees at Waze and merge it with its own map products. The company is reducing roles focused on Waze Ads monetization and providing employees with mobility resources and severance options. This move is part of Google’s efforts to reduce costs and comes after Alphabet announced job cuts across its product areas. Google also reported mixed results in advertising, with a decrease in advertising revenue growth but growth in revenue for its search advertising division.
PYMNTS discusses how H&M and Shein are diversifying their offerings by incorporating third-party brands into their platforms. H&M plans to enhance its selection of third-party brands both online and in physical stores, allowing customers to buy sneakers from brands like Adidas and New Balance along with clothing from Swedish mountaineering brand Klättermusen. This strategy aims to challenge online competitors such as Zalando, ASOS, and Shein. Similarly, Shein is collaborating with smaller-scale U.S. vendors to broaden its customer base in the United States and encourage customers to explore a wider range of offerings beyond clothing. PYMNTS highlights that the incorporation of third-party brands and marketplaces is a way for retailers to attract new customer segments, enhance the customer experience, and boost sales through cross-selling and upselling.
In terms of app downloads on the Shein platform, PYMNTS mentions that Temu, an online retailer owned by the parent company of Chinese social commerce giant Pinduoduo, reportedly surpassed Shein in mobile app downloads by November after entering the U.S. market in September. This indicates that Shein has experienced a deceleration in growth within its segment of affordable and fast-produced fashion, aligning with the broader trend in the fast-fashion industry.
In addition to PYMNTS own sources and reporting, the following third-party sources were consulted in the creation of this article: