Artificial intelligence (AI) is rapidly advancing, with the potential to automate a wide range of tasks initiated by just a one-sentence prompt. A recent survey of leading AI researchers suggests that by the year 2116, all human tasks could become highly automatable. This timeline has been moved up by 50 years from the original estimate of 2164, which was made just a year ago. Furthermore, there are developments emerging that will reshape the jobs of IT and business professionals within the next five to 10 years.
The survey, conducted by a team of authors from the University of California at Berkeley, aimed to measure the progress towards high-level machine intelligence. This level of intelligence is achieved when machines can accomplish every task better and more cost-effectively than human workers. The survey’s respondents provided predictions for when such intelligence will be feasible. In both 2022 and 2023, the predictions varied. However, the aggregate forecast for 2023 indicated a 50% chance of high-level machine intelligence by 2047, which is 13 years earlier than the 2060 date predicted in the 2022 survey.
While the long-range forecasts are interesting, let’s focus on what the survey suggests we can expect from AI in the next few years. According to the views of over 2,000 AI researchers, several capabilities could be just around the corner. Within the next five to 10 years, AI systems will be fully capable of delivering the following results, often in response to one-sentence prompts:
– Find and patch security flaws in open-source projects with over 100,000 users.
– Build a payment processing website from scratch, including frontend, backend, and secure payment integration.
– Provide phone banking services on par with human operators, including one-off tasks like ordering a replacement bank card.
– Write readable Python code for algorithms like quicksort, based on specifications and examples.
– Finetune large language models to improve their performance on predetermined benchmark metrics.
– Conduct machine-language studies and write papers that could be accepted at leading machine learning conferences.
The survey’s authors also highlight the impact of declining computing costs on AI development. If progress continues at its current pace, there is an estimated 10% chance that unaided machines will outperform humans in every possible task by 2027, and a 50% chance by 2047. This latter estimate is 13 years earlier than the prediction made in the 2022 survey.
The study also explores predictions related to the full automation of labor. The authors forecast a 10% chance of all human occupations becoming fully automatable by 2037, and a 50% chance by 2116. This is a significant shift from the previous prediction of 2164 in the 2022 survey.
Overall, AI is poised to revolutionize various industries and job roles. While the full automation of all human tasks may still be several decades away, the advancements expected within the next five to 10 years will undoubtedly reshape the way we work and interact with technology. As computing costs continue to decline, the potential for AI to surpass human capabilities becomes increasingly likely. It is crucial for businesses and professionals to adapt and embrace these changes to stay relevant in an AI-driven world.