Facing the chilly winter of downsizing, layoffs have become almost customary for overseas tech companies in recent years. Following Google, Amazon, Snap, and LinkedIn, music streaming service Spotify has initiated its third round of layoffs this year. The company revealed in an email that it would dismiss 1,500 employees, constituting 17% of its workforce, citing “rising costs.” This marks a total reduction of 25% of Spotify’s staff throughout the year.
Notably, Spotify’s CEO, Daniel Ek, stated in the related announcement that as the company streamlines its operations, it will be able to allocate more profits to the business. He emphasized that “streamlining does not mean reducing goals but implies achieving goals in a smarter and more efficient manner.” The so-called “smarter and more efficient” approach refers to artificial intelligence (AI). Spotify has chosen to use AI technology rather than human resources to enhance user experience and drive growth.
Earlier in February of this year, Spotify introduced an “AI DJ” feature based on AI technology to understand user preferences and provide recommended playlists, even predicting music, artists, or genres users might enjoy. Additionally, Spotify began using OpenAI’s Whisper voice translation tool in podcasts to translate selected English podcasts into Spanish, French, and German. In November, Spotify announced a collaboration with Google, restructuring its approach to recommending audiobooks and podcasts using Google Cloud’s Vertex AI Search language model.
In fact, it’s not just Spotify; Meta CEO Mark Zuckerberg informed employees after the latest round of layoffs that many jobs would not return as new technologies would enable the company to operate more efficiently. Similarly, IBM CEO Arvind Krishna mentioned that the company would temporarily freeze some hiring, possibly replacing 7,800 jobs with AI. Of the approximately 26,000 non-customer-facing positions at IBM, 30% are expected to be automated or replaced by AI tools within five years.
So, why is the change driven by AI happening so rapidly for these tech companies? In the current global consensus of reducing costs and increasing efficiency, AI, widely considered to enhance efficiency in certain areas, is conveniently used as an excuse for layoffs. While discussions about whether AI will replace workers in traditional industries are still ongoing, AI has already caused job losses for employees of tech companies.
In reality, the rapid adoption of AI by tech companies is primarily driven by cost considerations and the consequences of irrational high-speed expansions in the market environment. During the pandemic, stay-at-home policies in major Northern Hemisphere countries brought substantial new increments to the internet industry. Coupled with the continuous “watering” from the U.S. Federal Reserve, internet companies highly dependent on financing, from giants like Google, Microsoft, Amazon, Meta, to medium-sized enterprises like Spotify and Snap, seized this opportunity and engaged in a rapid expansion phase.
In an industry where “people” are the primary means of production, the expansion of tech companies often manifested as massive hiring. The result was that during the pandemic, almost all overseas internet companies were hiring based on the expectation that the internet would always be growing. However, after the pandemic, due to the non-existent imagined recovery and facing the stark reality, these companies must now pay the price for their past mistakes. Due to social responsibility constraints, it is not realistic to lay off employees all at once; hence, a new round of layoffs is being carried out under the banner of reducing costs and increasing efficiency.
At the same time, AIGC has made leaps and bounds in 2023. As entities at the forefront of the AI revolution, these companies, almost all of which are deeply involved in the AIGC wave, naturally discovered the feasibility of AI replacing human employees. However, in the beginning, AI was thought to “hollow out” routine jobs, replacing repetitive labor, such as intelligent customer service. This idea gained quick acceptance because similar transitions had occurred in human society before, as seen in the industrial revolution when machines replaced manual labor.
However, the reality is that creative work might be the first to be partially replaced. This is the core driving force behind decisions by companies like Spotify, Meta, and IBM to use AI to replace human workers. Even though it might be counterintuitive for creative work to be replaced by AI, as human decision-making is often influenced by emotions, morals, and ethics, AI relies on programs and algorithms for decision-making. This makes AI appear inadequate when handling complex ethical and moral issues, as perceived by the majority. However, those who have experienced GPT-4 might find that AI lacks emotion and empathy, a stereotype that is quickly dispelled.
Outstanding AI models such as GPT-4 and New Bing are particularly adept at empathizing with humans. Otherwise, why would AI chatbots integrated into various platforms using ChatGPT’s API become so popular? Therefore, AI chatbots actually need to be more friendly, patient, and empathetic than humans. Over the past year, AI models have gradually demonstrated their capabilities in writing, drawing, chatting, programming, and even understanding research papers. This proves that AI can not only handle repetitive tasks but also excel in creativity.
The logic that “because profession X requires trait Y, and trait Y is lacking in AI, people in profession X will not be unemployed” may no longer apply to mental work. Moreover, AI only needs to assume some tasks, enough to make most human positions in that profession obsolete. In the past, internet companies were hiring three people to do the work of five, paying four salaries. Now, these companies realize that with the assistance of AI, one person can do the work of five, and they only need to pay two salaries.
In summary, the development of AIGC in 2023 may have exceeded the expectations of most people. Outstanding AI models such as OpenAI’s GPT-4V and Google’s Gemini have greatly expanded the upper limit of AIGC. Consequently, Spotify is now bold enough to let AI, not human labor, drive the growth of its business. For the working class, this is undoubtedly not good news.