Skip to content
TheChatArchiveThe Chat Archive
Microsoftwizardlm-70b· MicrosoftHugging Face Arena Human Preference 55K

After AGI is deployed that is capable of research at human level, how long could it take…

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with vicuna-13b’s answer.
  1. User

    After AGI is deployed that is capable of research at human level, how long could it take to reach the level of scientific advancement similar to what humanity could've reached by the year 3000 if they didn't use AI? Estimate how many times faster AGIs could initially conduct research compared to humans, and what could be used to increase this serial speed difference, and what greater speedup could be achieved after the measures to increase it are implemented. Estimate how long it could've taken for humans to achieve this greater speedup of AGIs if they were to work towards it. Estimate how long it would take for AGIs themselves running at initial speedup to do the same. Estimate how long it would take to complete the project of reach the year 3000 level of advancement after the greater speedup is implemented. Estimate the total number of years the project would take, starting from the initial development of AGI capable of research at human level, proceeding to implement greater speedup while running at initial speedup, and finally reaching the target level of advancement while running at greater speedup. Be brief in the response.

  2. Microsoft Assistant

    It's challenging to provide precise estimates due to the abstract nature of the question and the numerous variables involved. However, I'll attempt to provide a rough approximation.

    1. Initial speed-up: AGIs could initially conduct research 10-100 times faster than humans, depending on their design, computational resources, and efficiency.
    2. Increasing the speed difference: This could be achieved by improving the AI's algorithms, expanding computational resources, and optimizing research workflows. A 10-fold increase might be achieved within a few years with significant investment.
    3. Greater speedup: With further advancements and investment, the AGI could potentially conduct research 1,000-10,000 times faster than humans.
    4. Time for humans to achieve greater speedup: It would likely take decades to a century for humans to achieve this speedup, given the limitations of human cognition and the pace of technological progress.
    5. Time for AGIs to achieve greater speedup: The AGI might be able to reach this speedup within a few years to a decade after its initial development, depending on the resources and efforts devoted to its improvement.
    6. Time to reach the target level of advancement: If the AGI is 1,000 times faster than humans, it might take around 300-3,000 years to reach the level of advancement that humanity would achieve by the year 3000 without AI.
    7. Total project time: Starting from the initial AGI development, implementing speedup measures, and finally reaching the target level of advancement, the total project time could be anywhere from 100 to 1,000 years, depending on the factors mentioned above.

    Please note that these estimates are highly speculative and should be taken as rough approximations. The actual timeline could be significantly shorter or longer, depending on various factors such as technological progress, societal and ethical considerations, and the rate of investment in AI research.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.