OpenAI removes the waitlist for DALL-E, giving everyone immediate access; OpenAI CEO Sam Altman says a public release is essential to develop the tech safely
None of these photos were taken by a camera. — All of these images were created by the artificial intelligence text …
Washington PostNitasha Tiku
Context & Ripple Effects
DALL-E moved from a research preview of DALL-E 2 to a credit-based beta over the preceding months. Removing the waitlist completes the shift from controlled admission to general user access.
Altman frames broad release as part of safety development, making access policy an operational feedback mechanism rather than only a distribution decision.
First-order effects
Prospective DALL-E users no longer need approval before trying the service, expanding OpenAI's active user base immediately.
OpenAI takes responsibility for observing and responding to safety issues across a much less curated population of users.
Second-order effects
The larger user base strengthens the commercial relevance of DALL-E's existing credit-based model, while putting pressure on OpenAI to make access controls work at consumer scale.
Developers gain a clearer signal that image generation is becoming a product platform, a trajectory later reflected in the public DALL-E API for app integration.
Third-order effects
If public access continues to be paired with safety iteration, frontier-model providers will increasingly treat staged availability and usage controls as core governance tools rather than temporary launch constraints.
Text-to-image services are moving toward a layered market: direct consumer access alongside paid, developer-facing distribution.
The trend: Generative-image AI is shifting from research-gated demos to commercially distributed platforms whose access rules double as safety governance.
Later today OpenAI will remove the waitlist for DALL-E. Anyone can sign up for immediate access. My look at the rapid spread of text-to-image generators & how AI companies write their own rules-w/public impact shaped by corporate ethics or terms of service https://t.co/XqZCiifDix
The accumulation of ethical debt will be immediate, while the mop-up of the resulting mess will go on in perpetuity. The way OpenAI rationalizes this ('we need to put it out into the world and learn') does not comport with its ‘benefit all of humanity’ mission. https://twitter.co…
“The introduction of DALL-E has triggered an explosion of text-to-image generators. Google & Meta quickly revealed that they had each been developing similar systems, but said their models weren't ready for the public.” https://www.washingtonpost.com/ ... by @nitashatiku
I'm glad to see this reporting from @nitashatiku about the rise of text-to-image systems and their dangers. She also collected quotes that range from laughable AI hype to alarming lack of responsibility. Here is a sample: https://www.washingtonpost.com/ ... >>
And danger! That's what was said of fiction. And photographs. And the telegraph. And radio. #moralpanic AI can now create any image in seconds, bringing wonder and danger https://www.washingtonpost.com/ ...
Here are my notes from trying DALL-E for the first time back in June. I still find it easiest to get good results from it than from Stable Diffusion or Midjourney, at least for straightforward text-to-image prompts https://simonwillison.net/... https://twitter.com/...
The Midjourney Discord has 2.7 million members. More than 1.5 million people joined DALL-E *in the beta*. Think we're still underestimating how widespread AI-generated images are going to be very, very soon. https://openai.com/...
I would be more worried about the disinformation potential of sophisticated AI-generated art if I had not spent a decade watching people believe mind-bogglingly stupid things based on the flimsiest and least convincing evidence you can imagine https://t.co/Ddl8pUNQkA
This is great, but how much of a factor was the emergence of a ton of other seemingly nearly as good, easily accessible models? https://twitter.com/...
“Once the line between truth and fake is eroded, everything will become fake. We will not be able to believe anything.” - @wamageed Dall-E access opens to everyone today: https://www.washingtonpost.com/ ...