--- next: - [[human-AI collaboration]] prev: - broader: - [[GPT-3]] publish: true aliases: - DALL-E uses GPT-3 to create images from text tags: - created: 2022-04-20T16:51:47-05:00 updated: 2022-06-20T10:55:13-05:00 --- » [[GPT-3]] + [[exponential technology|exponential tech]] --- DALL-E is a transformer language model that uses [[AI]] to create images from text. Other alternatives include [[VQGAN+CLIP]], [JAX-CLIP](https://colab.research.google.com/github/olaviinha/NeuralTextToImage/blob/main/JAX_CLIP_Guided_diffusion_nshepperd.ipynb#scrollTo=OoIL7ayzq7kC), [Disco Diffusion v5.2 with VR Mode](https://colab.research.google.com/github/alembics/disco-diffusion/blob/main/Disco_Diffusion.ipynb), among others. > [!warning] Given the prompt: 'a store front that has the word ‘openai’ written on it' ![https://cdn.openai.com/dall-e/v2/samples/text_rendering/e45259de88f6361db852504739eb9255_0.png](https://cdn.openai.com/dall-e/v2/samples/text_rendering/e45259de88f6361db852504739eb9255_0.png) > [!example] Architecture > <iframe id="reddit-embed" src="https://www.redditmedia.com/r/dalle2/comments/vdd86v/3d_render_of_a_very_futuristic_building_with_an/?ref_source=embed&amp;ref=share&amp;embed=true&amp;theme=dark" sandbox="allow-scripts allow-same-origin allow-popups" style="border: none;" teidentifier="zggy1m1bntsl4mx17wl" height="272" width="640" scrolling="no"></iframe> The terrifying thing about generating visual concepts from text how easy it'd be to [[deepfake]] a local news station website and weaponize it for [[rivalrous dynamics|rivalrous game theoretic purposes]]. > We recognize that work involving generative models has the potential for significant, broad societal impacts. In the future, we plan to analyze how models like DALL·E relate to societal issues like economic impact on certain work processes and professions, the potential for bias in the model outputs, and the longer term ethical challenges implied by this technology. # Refs - [DALL·E: Creating Images from Text](https://openai.com/blog/dall-e/) - via:: [[@Eric Schmidt]] on [[Making Sense with Sam Harris]] > [!cite] [The AI that creates any picture you want, explained](https://www.youtube.com/watch?v=SVcsDDABEkM) from [[vox]] > <iframe width="1280" height="720" src="https://www.youtube.com/embed/SVcsDDABEkM" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe> # Related