OpenAI News & Discussions

[firestar464]

I agree, but I also need to caution people against black-and-white thinking. Hell, it's entirely possible that Ilya and the others have reconciled in a way. It's even possible that your average internet user hates Ilya more than Sam himself does

No one knows anything bruh

[Cyber_Rebel]

Concerning the possible Open AI breakthrough, I often feel like AI Explained's channel is a bit more nuanced than most:

[firestar464]

Nature's word on what happened at OpenAI:

https://www.nature.com/articles/d41586-023-03700-4

[raklian]

Nature's word on what happened at OpenAI:

https://www.nature.com/articles/d41586-023-03700-4

Unfettered capitalism galore.

[spryfusion]

[wjfox]

[funkervogt]

Humans say ChatGPT's answers to life questions are more empathetic than those generated by professional human advice columnists.
https://www.frontiersin.org/articles/10 ... 3.1281255/

[raklian]

Humans say ChatGPT's answers to life questions are more empathetic than those generated by professional human advice columnists.
https://www.frontiersin.org/articles/10 ... 3.1281255/

Ironic.

[weatheriscool]

Microsoft to Upgrade Copilot With GPT-4 Turbo and Improved Dall-E 3
The upgrades should be rolling out soon, making it the only free way to get access to OpenAI's latest AI model.
By Josh Norem December 6, 2023

As the AI race heats up, companies must keep their foot on the proverbial gas pedal to ensure they don't fall behind. Microsoft seems to be doing precisely that by announcing it will upgrade the AI model in its Copilot assistant to OpenAI's latest GPT-4 Turbo. It has also announced improvements to the Dall-E 3 image generator and several new features designed to improve Copilot's usefulness.

https://www.extremetech.com/computing/m ... d-dall-e-3

[weatheriscool]

OpenAI Revenues $1.5B/Year and Growth Path to $100 Billion
December 13, 2023 by Brian Wang

https://www.nextbigfuture.com/2023/12/o ... llion.html

Kelvin Mu, Translink Capital, estimates that OpenAI’s $1.5 billion in revenue is split pretty evenly right now between B2C (GPT plus subscriptions) and B2B (API tokens).

B2B will be the larger future revenue.

For B2C:
The total TAM is about $10-20B. Tech incumbents make more now. Google ($280B), Microsoft ($218B), and Meta ($117B).

The breakeven for the $20/month GPT+ subscription fee is about 50 pages of text output per day at the current token cost. There are currently about 3 million customers paying $240 per year for about $700 million in revenue.

Nextbigfuture notes the token and other costs could decrease with more efficient future AI. The B2C TAM could increase if AI is able to deliver more future customer value.

For B2B:
The total TAM at $200B+ under current token pricing. This assumes an addressable population of 2 billion and an ARPU of at least $100.