4 ms·

Ask HN: Bet on the Success of OpenAI/ChatGPT

6 points by smithcoin 4y ago

A friend of mine and I were debating the likely "disruptions" that ChatGPT will have in the coming years and how it will alter various industries. The timeline he suggested was 2 years. The questions I have are this: - What do you think is a reasonable litmus-test for measuring this? (e.g number of employed workers in AI) - He was willing to and I quote "bet my career" on the outcome of this. I have no need for his career. In the spirit of the bet what would you think is a prize that is worthy of this? (e.g something tongue in cheek like dinner at an AI powered White Castle[0]) I will be posting the bet on Longbets.org. I have attempted to leave my position/bias out of this post in relation to the bet. [0] https://www.engadget.com/white-castle-flippy-2-100-locations-205023731.html

OhNoNotAgain_99 4y ago
jerrygenser 4y ago
It's very difficult to measure hard numbers but a general framework would be ROI. How much has been invested in hunan-years worked and CO2 emissions training the models vs. how much value is actually created from the resulting products and services. Any net positive I would consider success.
ddmma 4y ago
Too bad we cannot embrace the humanity outcomes due to ancient unsolved legacies such as war and poverty. This gives more hope and boost on further development.
dbieber 4y ago
> e.g something tongue in cheek like dinner at an AI powered White Castle How about robot assembled salad from Sweetgreen [1]? I'd suggest autonomous pizza but Zume is refocusing [2]. Perhaps all topped with basil never touch by human hands [3]? [1] https://techcrunch.com/2021/08/25/salad-chain-sweetgreen-buys-kitchen-robotics-startup-spyce/ https://techcrunch.com/2021/08/25/salad-chain-sweetgreen-buy... [2] https://www.restaurantbusinessonline.com/technology/zume-shuts-down-its-robot-powered-pizza-business https://www.restaurantbusinessonline.com/technology/zume-shu... [3] https://ironox.com/ https://ironox.com/
f0e4c2f7 4y ago
Best metric I can think of is Revenue on the ChatGpt api. Not sure if that will be public though. Maybe some vague public number? Could bet on over or under $50M/yr, $1B/yr? Prize could be an A100 or whatever replaces it in 2 years. Also if your friend is willing to "bet their career on it" there is actually a place where they can literally do that, by applying to OpenAI.

6 points by smithcoin 4y ago

OhNoNotAgain_99 4y ago

jerrygenser 4y ago

It's very difficult to measure hard numbers but a general framework would be ROI. How much has been invested in hunan-years worked and CO2 emissions training the models vs. how much value is actually created from the resulting products and services. Any net positive I would consider success.

ddmma 4y ago

Too bad we cannot embrace the humanity outcomes due to ancient unsolved legacies such as war and poverty. This gives more hope and boost on further development.

dbieber 4y ago

> e.g something tongue in cheek like dinner at an AI powered White Castle How about robot assembled salad from Sweetgreen [1]? I'd suggest autonomous pizza but Zume is refocusing [2]. Perhaps all topped with basil never touch by human hands [3]? [1] https://techcrunch.com/2021/08/25/salad-chain-sweetgreen-buys-kitchen-robotics-startup-spyce/ https://techcrunch.com/2021/08/25/salad-chain-sweetgreen-buy... [2] https://www.restaurantbusinessonline.com/technology/zume-shuts-down-its-robot-powered-pizza-business https://www.restaurantbusinessonline.com/technology/zume-shu... [3] https://ironox.com/ https://ironox.com/

f0e4c2f7 4y ago

Best metric I can think of is Revenue on the ChatGpt api. Not sure if that will be public though. Maybe some vague public number? Could bet on over or under $50M/yr, $1B/yr? Prize could be an A100 or whatever replaces it in 2 years. Also if your friend is willing to "bet their career on it" there is actually a place where they can literally do that, by applying to OpenAI.