Sam Altman writes that DALL•E 2 can create and edit images from natural language

Sam Altman writes that DALL•E 2 can create and edit images from natural language

On April 6, 2022, Sam Altman described DALL•E 2 as a research launch: an AI tool that can create and edit images from natural language instructions. He called it useful and creativity-enhancing, and argued that it points to a broader computer interface trend in which people state what they want and computers carry it out. Altman said DALL•E 2 could create a “complete work,” unlike a coding copilot that he said was still far from creating a full program. He suggested this might be an early example of AI’s impact on labor markets. In his view, AI could create jobs and improve existing ones by handling boring tasks, while also making some jobs less relevant. These are his expectations, not reported labor-market outcomes. Altman also said predictions about AI’s effects on work are difficult, and described the model as powerful enough to raise potential downsides. He said the research launch was intended to help researchers and artists work on reducing those downsides and to give people time to adapt. He hoped for a product launch in the summer, but this post does not establish whether or when that happened. What safeguards or access conditions would shape a later product launch?

Publication: 8e3b7507-75bb-f6cc-eb60-17c2dcfc6bcf

attribution — pass: The post is on Sam Altman's own blog and is written in his voice. The draft correctly attributes the description of DALL•E 2, the interface-trend argument, the labor-market views and the launch rationale to Altman, framing them as his views and expectations.

Today we did a research launch of DALL•E 2, a new AI tool that can create and edit images from natural language instructions.

summary — pass: The title, summary and question accurately reflect the post's main points: the research launch, the natural-language interface trend, the contrast with Copilot, the labor-market argument, the downsides and the incremental deployment rationale. Narrowing the difficulty of AI predictions to effects on work matches the source's own example. The draft is understandable to a non-specialist reader, and the question follows from the post's discussion of downsides.

It’s a reminder that predictions about AI are very difficult to make.

date — pass: The originalPublished instant, 2022-04-06T18:15:13Z, matches the stated date of April 6, 2022 at day precision. The hoped-for summer product launch is presented as Altman's hope, with no claim about its outcome, and there is no hindsight framing. The evidence kind is publisher-page, so no archive disclosure is required.

Hopefully this summer, we’ll do a product launch and people will be able to use it for all sorts of things.

support — pass: Each claim is supported by a verbatim excerpt and labeled as source-stated. The draft explicitly marks the job effects as Altman's expectations rather than outcomes. The question about safeguards and access conditions does not assert any unsupported facts.

I think it’s important to be honest that it’s increasingly going to make some jobs not very relevant (like technology frequently does).

media — pass: No media was selected, and no verified media candidates were supplied, so there is no media identity or relevance problem.