Trick for better AI predictions, Humane AI pin slammed, Atlas robot: AI Eye

All the latest insights into the rapidly advancing field of AI.

Highlights: Featured Video — Atlas Robot, The Controversy Surrounding Humane’s AI pin, AI’s role in preserving memories of historical tragedies, Impact of mid-level AI on knowledge retention, The value of AI services, Can individuals without coding knowledge utilize AI? Latest Relevant AI Updates.

Harnessing the Past to Predict the Future

New methods have emerged to encourage ChatGPT to engage in its least preferred task: foreseeing future events.

Recent studies propose that to achieve accurate predictions from ChatGPT, prompting it to craft a narrative set in the future, reflecting on yet-to-occur events, is the most effective approach.

The study encompassed an evaluation of 100 diverse prompts, categorized between direct forecasts (e.g., anticipating the winner of the 2022 Oscars for Best Actor) versus “future narratives,” such as instructing the chatbot to compose a story about a family watching the 2022 Oscars on television and portraying the scenario as the announcement of the best actor award unfolds.

The narrative technique yielded more precise outcomes. In a similar vein, deriving an insightful prediction on interest rates involved tasking the model to construct a story centering around Fed Chair Jerome Powell reflecting on past occurrences. Reddit users experimented with this method, resulting in a forecast of an interest rate increase in June and a financial crisis in 2030.

In theory, directing ChatGPT to construct a Cointelegraph news piece set in 2025, revisiting this year’s significant Bitcoin price fluctuations, should yield a more accurate price projection than a direct inquiry.

However, the research faces two potential constraints: the choice of the 2022 Oscars by the researchers, informed by their knowledge of the winners, whereas ChatGPT’s training data only extended to September 2021. Nonetheless, there are several instances of ChatGPT providing insights it technically shouldn’t possess from its training data.

Another challenge lies in the suspicion that OpenAI may have purposely disrupted ChatGPT’s predictive capabilities, implying that this new strategy…

..

Expand Insights

Recommended For You

Leave a Reply

Your email address will not be published. Required fields are marked *

Protected by Security by CleanTalk