Training Data News

For at least one day, the model was generally available to anyone in the United States willing to pay enough without waiting lists or "beta" programs

It is discussed during a segment of "HatGPT"

By default, the generated videos will be watermarked

A competing model has challenged Sora's dominance in the text-to-video space

A poll of Manifold users will say that 20% or less have made a Sora video in the last month

OpenAI will be sued over the model

It was trained on data created in a physics/game engine (eg Unreal Engine)

A post claiming Sora video is real will go viral with > 1 million engagements

It will be noticeably worse at or largely unable to generate fast-paced animation

A second major version of the model has been released

A video produced by the model has been used for widely spread misinformation, as reported by a major news outlet

It has been referenced in a legal case about deepfakes

costs for an average 1 minute HD (or higher quality) video will be lower than $0.50

Video will include sound

A major studio will use this in a movie or tv show

It can be used during conversations with ChatGPT on the OpenAI website

It prompts the Hard Fork Podcast to rant about AI model names

A youtube video made only with Sora will get > 100M views

It will be pay-per-use (or credit based) as opposed to as part of a monthly subscription

It will be jailbroken to make a porn video

it'll be legaly banned in at least one EU country

A youtube movie >2h will be made with only Sora and splicing videos together will get > 10M views

At least 2 Manifold questions will contain a Sora-generated video in their header

Sora will be part of GPT model

It will be the most popular text to video tool (determined by google search trends)

Public access was revoked after release, even if it is later restored

There will be a new monthly subscription that includes sora and dalle for creatives

It has been integrated as a feature on a major social media platform

It can generate videos over 10 minutes long

It has a logo separate from the OpenAI logo

It will be the SOTA for text to video

It has been renamed

It will be free to use

A poll of Manifold users will say that 30% or more have created a Sora video in the last month

The model has had a non-trivial effect on the everday life of the average American, as judged by @Bayesian

It can create a fully coherent short film from a prompt (20-40 minutes)

The Sora line of models proves to be useful for purposes where the video is secondary, such as research into physics, medicine, and weather

A third major version of the model has been released

OpenAI will release the number of model parameters

Full description of model architecture will be public

Eliezer Yudkowsky states or implies that future versions of the Sora line of models - specifically, by name - are an existential threat to civilization

The Sora line of models are being used as simulators for legal investigations - including but not limited to predicting events leading to accidents and crimes

A version of the model was the cause of the YES resolution of the "Weak AGI" Metaculus market

OpenAI will lose a lawsuit over the model

Eliezer Yudkowsky has stated or implied that the current version or an obsolete version of the model poses or had posed an existential threat to civilization

A nyt bestselling author will release their own bestseller movie/tv adaptation using sora

YES

The case is fully resolved somehow, including all appeals, by the end of 2026.

The case is fully resolved somehow, including all appeals, by the end of 2025.

Disney sues OpenAI for copyright infringement before the NYT case concludes.

NYT and OpenAI announce a partnership that goes beyond access to training data, such as the one OpenAI made with Politico.

The case settles for more than $10 million before a verdict is reached.

The case settles before a verdict is reached.

New York Times wins at least $1 in damages via a verdict.

The case is fully resolved somehow, including all appeals, by the end of 2024.

The case appears before the Supreme Court.

Nintendo sues OpenAI for copyright infringement before the NYT case concludes.

OpenAI wins via an outright verdict in their favor.

The case settles for more than $100 million before a verdict is reached.

New York Times wins at least $10 million in damages via a verdict.

New York Times wins at least $100 million in damages via a verdict.

The Supreme Court rules in favor of The New York Times and upholds damages and compensation with a net present value of $10 million or more, but does not order GPT-4 deleted.

New York Times wins at least $1 billion in damages via a verdict.

New York Times wins a verdict ordering GPT-4 to be deleted.

The Supreme Court orders GPT-4 to be deleted, or OpenAI otherwise agrees to delete GPT-4 on the basis of this case.

Runs entirely on silicon-based non-quantum machines

The existence of the AGI and the model name were leaked prior to release and official announcement

Developed by a US company

Trained with synthetic data and self-play

Transformer based or mix of Transformer and other architectures

Doomsday clock moved at least one full minute closer to midnight due to this AI

Has an internal World Model able to simulate basic physics without external tools

No consensus that it's AGI until over a month after being announced and interacting with people outside the lab.

Not available in EU countries initially for the first 4 months after public release

Based on/inspired by algorithms from the human brain (beyond neural networks).

Public announcement was made after 1 year or more of finishing training

Infinite context window

Solves at least one of the seven Millennium Problems

Not a static model; weights change during inference

Used by a Figure humanoid robot

Is OpenAI's GPT-6, whatever name or architecture it has

Based on/inspired by OpenAI's Sora (resolves YES if explicitly stated by developers in announcement or paper)

This market will resolve N/A

Developed by a Chinese company

Open sourced

Is OpenAI's GPT-5, whatever name or architecture it has

Developed by the open-source community instead of a company

YES

NO

NO

2025

2026

2030

2024

2028

2029

2027

NO

NO

YES

NO