In the near future, the capability to program artificial intelligence (AI) to perform a wide range of tasks based on a single-sentence prompt is on the horizon. A recent survey conducted among prominent AI researchers suggests that this advancement will have a significant impact on job roles in the coming years. There is a high likelihood that by 2116, the majority of routine tasks will be automated, marking a substantial shift in the employment landscape.
It is important to note that the timeline for this development has been extended by 50 years from the initial projection of 2164, which was predicted just a year ago. Despite the seemingly distant timeframe, rapid innovations expected within the next five to ten years are poised to transform the IT and business sectors.
Moreover, the integration of conceptual AI underscores the need for developers to expand their skills across various domains to stay relevant in the evolving technological landscape.
These insights are derived from a survey involving 2,778 AI researchers, shedding light on the potential of AI to enhance skill sets in both the short and long term.
The survey’s main goal was to evaluate progress towards achieving high-level system intelligence, where untrained machines can efficiently outperform human workers in all tasks. This assessment, led by researchers from the University of California at Berkeley, including Katja Grace and Harlan Stewart, generated a range of predictions regarding the feasibility of such intelligence by 2022 and 2023. Notably, the collective forecast for 2023 anticipates a 50% chance of high-level machine intelligence by 2047, which is 13 years earlier than the 2060 estimate from the survey.
While the current focus is on short-term advancements, the study also outlines the potential impacts of AI in the years to come. Insights from over 2,000 AI researchers suggest several upcoming advancements in the field.
These forthcoming capabilities, often triggered by concise prompts, are expected to be fully automated by AI systems within the next five to ten years. Even the sophisticated language models (LLMs) supporting these functions will be enhanced by AI technology:
- Detecting and addressing security vulnerabilities in open-source projects with a large user base based on a single prompt.
- Independently creating a payment processing website, including payment processing, frontend and backend operations, and secure payment integration.
- Providing mobile banking solutions alongside human operators for specialized tasks like assisting customers with card replacements or guiding them through website navigation.
- Generating concise and readable Python code for implementing algorithms such as quicksort based on specific requirements and scenarios.
- Improving existing open-source LLMs based on brief prompts to enhance their performance against predefined benchmarks.
- Engaging in language research and producing papers on system-related topics, addressing experimental questions in machine learning autonomously.
Experts highlight that the decreasing costs of technology will drive significant progress in AI. The authors of the study predict that if AI development continues at its current pace, machines surpassing humans in all activities independently could have a 10% likelihood by 2027 and a 50% likelihood by 2047, reaching this milestone 13 years earlier than previously thought.
Additionally, the study identifies the top five AI devices relevant to the workplace.
The authors also discuss the concept of “full work technology” in their analysis. In comparison to the 2022 assessment, the probability of automating all human tasks entirely is estimated to be 10% by 2037 and could potentially reach 50% as early as 2116.
