DALL-E

{{Short description|Image-generating deep learning model}} {{Use dmy dates|date=June 2022}} {{Infobox software | name = DALL-E | logo = DALL-E 2 Signature.svg | logo caption = Watermark present on DALL-E images | screenshot = DALL-E 2 artificial intelligence digital image generated photo.jpg | caption = An image generated by DALL-E 2, from the prompt "Teddy bears working on new AI research underwater with 1990s technology" | author = | developer = OpenAI | released = {{start date and age|df=y|2021|1|5}} | latest release version = 3 | latest release date = {{start date and age|df=y|2023|8|10}} | replaced_by = GPT Image | repo = | programming language = | operating system = | genre = Text-to-image model | platform = Cloud computing platforms | license = Proprietary service }} {{OpenAI series}} {{Artificial intelligence}}

'''DALL-E''', '''DALL-E 2''', and '''DALL-E 3''' (stylised '''DALL·E''') are text-to-image models developed by OpenAI using deep learning methodologies to generate digital images from natural language descriptions known as ''prompts''.

The first version of DALL-E was announced in January 2021. In the following year, its successor DALL-E 2 was released. DALL-E 3 was released natively into ChatGPT for ChatGPT Plus and ChatGPT Enterprise customers in October 2023,<ref name="David-2023" /> with availability via OpenAI's API<ref name="platform.openai.com" /> and "Labs" platform provided in early November.<ref name="Niles-2023" /> Microsoft implemented the model in Bing's Image Creator tool and plans to implement it into their Designer app.<ref name="Mehdi-2023" /> With Bing's Image Creator tool, Microsoft Copilot runs on DALL-E 3.<ref name="CopilotDalle3">{{cite web |title=AI art improvements with DALL-E 3 |url=https://www.microsoft.com/en-us/microsoft-copilot/for-individuals/do-more-with-ai/ai-art-and-creativity/image-creator-improvements-dall-e-3?form=MA13KP |website=Microsoft |access-date=October 1, 2024}}</ref> In March 2025, DALL-E-3 was replaced in ChatGPT by GPT Image's native image-generation capabilities.<ref>{{Cite web |last1=Zeff |first1=Kyle |last2=Wiggers |first2=Maxwell |date=2025-03-25 |title=ChatGPT's image-generation feature gets an upgrade |url=https://techcrunch.com/2025/03/25/chatgpts-image-generation-feature-gets-an-upgrade/ |access-date=2025-05-12 |website=TechCrunch |language=en-US}}</ref>

== History and background == DALL-E was revealed by OpenAI in a blog post on 5 January 2021, and uses a version of GPT-3 modified to generate images.<ref name="vb" />

On 6 April 2022, OpenAI announced DALL-E 2, a successor designed to generate more realistic images at higher resolutions that "can combine concepts, attributes, and styles".<ref name="OpenAI-2" /> On 20 July 2022, DALL-E 2 entered into a beta phase with invitations sent to 1 million waitlisted individuals;<ref name="OpenAI-2022b" /> users could generate a certain number of images for free every month and may purchase more.<ref name="Allyn-2022" /> Access had previously been restricted to pre-selected users for a research preview due to concerns about ethics and safety.<ref name="labs.openai.com" /><ref name="Guardian-2022" /> On 28 September 2022, DALL-E 2 was opened to everyone and the waitlist requirement was removed.<ref name="OpenAI-2022c" /> In September 2023, OpenAI announced their latest image model, DALL-E 3, capable of understanding "significantly more nuance and detail" than previous iterations.<ref name="OpenAI" /> In early November 2022, OpenAI released DALL-E 2 as an API, allowing developers to integrate the model into their own applications. Microsoft unveiled their implementation of DALL-E 2 in their Designer app and Image Creator tool included in Bing and Microsoft Edge.<ref name="OpenAI-2022d" /> The API operates on a cost-per-image basis, with prices varying depending on image resolution. Volume discounts are available to companies working with OpenAI's enterprise team.<ref name="Wiggers-2022" />

The software's name is a portmanteau of the names of animated robot Pixar character WALL-E and the Spanish surrealist artist Salvador Dalí.<ref name="tc" /><ref name="vb" />

In February 2024, OpenAI began adding watermarks to DALL-E generated images, containing metadata in the C2PA (Coalition for Content Provenance and Authenticity) standard promoted by the Content Authenticity Initiative.<ref>{{Cite web |last=Growcoot |first=Matt |date=2024-02-08 |title=AI Images Generated on DALL-E Now Contain the Content Authenticity Tag |url=https://petapixel.com/2024/02/08/ai-images-generated-on-dall-e-now-contain-the-content-authenticity-tag/ |access-date=2024-04-04 |website=PetaPixel |language=en}}</ref>

== Technology == The first generative pre-trained transformer (GPT) model was initially developed by OpenAI in 2018,<ref name="Radford-2018" /> using a Transformer architecture. The first iteration, GPT-1,<ref name="GPT-2023" /> was scaled up to produce GPT-2 in 2019;<ref name="Radford-2019" /> in 2020, it was scaled up again to produce GPT-3, with 175 billion parameters.<ref name="Brown-2020" /><ref name="vb" /><ref name="dallepaper" />

=== DALL-E === DALL-E has three components: a discrete VAE, an autoregressive decoder-only Transformer model (12 billion parameters) similar to GPT-3, and a CLIP pair of image encoder and text encoder.<ref name="Ramesh-2022" />

The discrete VAE can convert an image to a sequence of tokens, and conversely, convert a sequence of tokens back to an image. This is necessary as the Transformer model does not directly process image data.<ref name="Ramesh-2022" />

The input to the Transformer model is a sequence of tokenised image caption followed by tokenised image patches. The image caption is in English, tokenised by byte pair encoding (vocabulary size 16384), and can be up to 256 tokens long. Each image is a 256×256 RGB image, divided into 32×32 patches of 4×4 each. Each patch is then converted by a discrete variational autoencoder to a token (vocabulary size 8192).<ref name="Ramesh-2022" />

DALL-E was developed and announced to the public in conjunction with CLIP (Contrastive Language-Image Pre-training).<ref name="Heaven-2021" /> CLIP is a separate model based on contrastive learning that was trained on 400 million pairs of images with text captions scraped from the Internet. Its role is to "understand and rank" DALL-E's output by predicting which caption from a list of 32,768 captions randomly selected from the dataset (of which one was the correct answer) is most appropriate for an image.<ref>{{Cite conference |last1=Radford |first1=Alec |last2=Kim |first2=Jong Wook |last3=Hallacy |first3=Chris |last4=Ramesh |first4=Aditya |last5=Goh |first5=Gabriel |last6=Agarwal |first6=Sandhini |last7=Sastry |first7=Girish |last8=Askell |first8=Amanda |last9=Mishkin |first9=Pamela |display-authors=3 |date=2021-07-01 |title=Learning Transferable Visual Models From Natural Language Supervision |url=https://proceedings.mlr.press/v139/radford21a |conference=Proceedings of the 38th International Conference on Machine Learning |language=en |publisher=PMLR |pages=8748–8763 |last10=Clark |first10=Jack |last11=Krueger |first11=Gretchen |last12=Sutskever |first12=Ilya}}</ref>

A trained CLIP pair is used to filter a larger initial list of images generated by DALL-E to select the image that is closest to the text prompt.<ref name="Ramesh-2022" />

=== DALL-E 2 === DALL-E 2 uses 3.5 billion parameters, a smaller number than its predecessor.<ref name="Ramesh-2022" /> Instead of an autoregressive Transformer, DALL-E 2 uses a diffusion model conditioned on CLIP image embeddings, which, during inference, are generated from CLIP text embeddings by a prior model.<ref name="Ramesh-2022" /> This is the same architecture as that of Stable Diffusion, released a few months later.

=== DALL-E 3 === While a technical report was written for DALL-E 3, it does not include training or implementation details of the model, instead focusing on the improved prompt following capabilities developed for DALL-E 3.<ref>{{cite arXiv |eprint=2006.11807 |last1=Shi |first1=Zhan |last2=Zhou |first2=Xu |last3=Qiu |first3=Xipeng |last4=Zhu |first4=Xiaodan |title=Improving Image Captioning with Better Use of Captions |date=2020 |class=cs.CV }}</ref>

== Capabilities == DALL-E can generate imagery in multiple styles, including photorealistic imagery, paintings, and emoji.<ref name="vb" /> It can "manipulate and rearrange" objects in its images,<ref name="vb" /> and can correctly place design elements in novel compositions without explicit instruction. Thom Dunn writing for ''BoingBoing'' remarked that "For example, when asked to draw a daikon radish blowing its nose, sipping a latte, or riding a unicycle, DALL-E often draws the handkerchief, hands, and feet in plausible locations."<ref name="boing" /> DALL-E showed the ability to "fill in the blanks" to infer appropriate details without specific prompts, such as adding Christmas imagery to prompts commonly associated with the celebration,<ref name="extreme" /> and appropriately placed shadows to images that did not mention them.<ref name="engadget" /> Furthermore, DALL-E exhibits a broad understanding of visual and design trends.{{Citation needed|date=July 2022}}

DALL-E can produce images for a wide variety of arbitrary descriptions from various viewpoints<ref name="Marcus-2022" /> with only rare failures.<ref name="tc" /> Mark Riedl, an associate professor at the Georgia Tech School of Interactive Computing, found that DALL-E could blend concepts (described as a key element of human creativity).<ref name="cnbc" /><ref name="bbc" />

Its visual reasoning ability is sufficient to solve Raven's Matrices (visual tests often administered to humans to measure intelligence).<ref name="dale" /><ref name="OpenAI-2021" />

thumb|An accurate image generated by DALL-E 3 based on the text prompt "An illustration of an avocado sitting in a therapist's chair, saying 'I just feel so empty inside' with a pit-sized hole in its centre. The therapist, a spoon, scribbles notes" DALL-E 3 follows complex prompts with more accuracy and detail than its predecessors, and is able to generate more coherent and accurate text.<ref name="Edwards-2023" /><ref name="OpenAI" /> DALL-E 3 is integrated into ChatGPT Plus.<ref name="OpenAI" />

=== Image modification === {{multiple image | align = | width = 130 | image1 = DALL-E 2 variation 1.png | image2 = DALL-E 2 variation 2.png | footer = Two "variations" of ''Girl With a Pearl Earring'' generated with DALL-E 2 }} Given an existing image, DALL-E 2 and DALL-E 3 can produce "variations" of the image as individual outputs based on the original, as well as edit the image to modify or expand upon it. The "inpainting" and "outpainting" abilities of these models use context from an image to fill in missing areas using a medium consistent with the original, following a given prompt.

For example, this can be used to insert a new subject into an image, or expand an image beyond its original borders.<ref name="Coldewey-2022" /> According to OpenAI, "Outpainting takes into account the image’s existing visual elements — including shadows, reflections, and textures — to maintain the context of the original image."<ref name="OpenAI-2022" />

=== Technical limitations === DALL-E 2's language understanding has limits. It is sometimes unable to distinguish "A yellow book and a red vase" from "A red book and a yellow vase" or "A panda making latte art" from "Latte art of a panda".<ref name="Saharia-2022" /> It generates images of an astronaut riding a horse when presented with the prompt "a horse riding an astronaut".<ref name="Marcus-2022a" /> It also fails to generate the correct images in a variety of circumstances. Requesting more than three objects, negation, numbers, and connected sentences may result in mistakes, and object features may appear on the wrong object.<ref name="Marcus-2022" /> Additional limitations include generating text, ambigrams and other forms of typography, which often results in dream-like gibberish. The model also has a limited capacity to address scientific information, such as astronomy or medical imagery.<ref name="Strickland-2022" /> [[File:AI_tanuki_Japanese_text.jpg|thumb|An attempt to generate Japanese text using the prompt "a person pointing at a tanuki, with a speech bubble that says '{{lang|ja|これは狸です！}}{{' "}}, which results in the text being rendered with nonsensical kanji and kana]]

== Ethical concerns == DALL-E 2's reliance on public datasets influences its results and leads to algorithmic bias in some cases, such as generating higher numbers of men than women for requests that do not mention gender.<ref name="Strickland-2022" /> DALL-E 2's training data was filtered to remove violent and sexual imagery, but this was found to increase bias in some cases such as reducing the frequency of women being generated.<ref name="OpenAI-2022a" /> OpenAI hypothesise that this may be because women were more likely to be sexualised in training data which caused the filter to influence results.<ref name="OpenAI-2022a" /> In September 2022, OpenAI confirmed to ''The Verge'' that DALL-E invisibly inserts phrases into user prompts to address bias in results; for instance, "black man" and "Asian woman" are inserted into prompts that do not specify gender or race.<ref name="Vincent-2022" /> OpenAI claims to address concerns for potential "racy content"{{snd}}containing nudity or sexual content generation, with DALL-E 3 through input/output filters, blocklists, ChatGPT refusals, and model level interventions.<ref name="DALLE3SystemCard">{{cite journal |last1=OpenAI |title=DALL-E 3 System Card |date=October 3, 2023 |url=https://cdn.openai.com/papers/DALL_E_3_System_Card.pdf}}</ref> However, DALL-E 3 continues to disproportionally represent people as White, female, and youthful. Users are able to somewhat remedy this through more specific prompts for image generation.

A concern about DALL-E 2 and similar image generation models is that they could be used to propagate deepfakes and other forms of misinformation.<ref name="Taylor" /><ref name="Knight-2022" /> As an attempt to mitigate this, the software rejects prompts involving public figures and uploads containing human faces.<ref name="vice" /> Prompts containing potentially objectionable content are blocked, and uploaded images are analysed to detect offensive material.<ref name="docs" /> A disadvantage of prompt-based filtering is that it is easy to bypass using alternative phrases that result in a similar output. For example, the word "blood" is filtered, but "ketchup" and "red liquid" are not.<ref name="Lane-2022" /><ref name="docs" />

Another concern about DALL-E 2 and similar models is that they could cause technological unemployment for artists, photographers, and graphic designers due to their accuracy and popularity.<ref name="Goldman-2022" /><ref name="Blain-2022" /> DALL-E 3 is designed to block users from generating art in the style of currently-living artists.<ref name="OpenAI" /> While OpenAI states that images produced using these models do not require permission to reprint, sell, or merchandise,<ref>{{Cite web |title=DALL·E 3 |url=https://openai.com/index/dall-e-3/ |access-date=2025-04-10 |website=openai.com |language=en-US}}</ref> legal concerns have been raised regarding who owns those images.<ref>{{Cite web |last=centerforartlaw |date=2022-11-21 |title=Art-istic or Art-ificial? Ownership and copyright concerns in AI-generated artwork - Center for Art Law |url=https://itsartlaw.org/2022/11/21/artistic-or-artificial-ai/ |access-date=2025-04-10 |website=itsartlaw.org |language=en-US}}</ref><ref>{{Cite news |date=2023-04-07 |title=Generative AI Has an Intellectual Property Problem |url=https://hbr.org/2023/04/generative-ai-has-an-intellectual-property-problem |access-date=2025-04-10 |work=Harvard Business Review |language=en |issn=0017-8012}}</ref>

In 2023 Microsoft pitched the United States Department of Defense to use DALL-E models to train battlefield management systems.<ref>{{cite news |last1=Biddle |first1=Sam |title=Microsoft Pitched OpenAI's DALL-E as Battlefield Tool for U.S. Military |url=https://theintercept.com/2024/04/10/microsoft-openai-dalle-ai-military-use/ |work=The Intercept |date=10 April 2024}}</ref> In January 2024 OpenAI removed its blanket ban on military and warfare use from its usage policies.<ref>{{cite news |last1=Biddle |first1=Sam |title=OpenAI Quietly Deletes Ban on Using ChatGPT for "Military and Warfare" |url=https://theintercept.com/2024/01/12/open-ai-military-ban-chatgpt/ |work=The Intercept |date=12 January 2024}}</ref>

== Reception == thumb|upright=1.35|Images generated by DALL-E upon the prompt: "an illustration of a baby daikon radish in a tutu walking a dog" Most coverage of DALL-E focuses on a small subset of "surreal"<ref name="Heaven-2021" /> or "quirky"<ref name="cnbc" /> outputs. DALL-E's output for "an illustration of a baby daikon radish in a tutu walking a dog" was mentioned in pieces from ''Input'',<ref name="input" /> NBC,<ref name="nbc" /> ''Nature'',<ref name="nature" /> and other publications.<ref name="vb" /><ref name="Knight-2021" /><ref name="cnn" /> Its output for "an armchair in the shape of an avocado" was also widely covered.<ref name="Heaven-2021" /><ref name="bbc" />

''ExtremeTech'' stated "you can ask DALL-E for a picture of a phone or vacuum cleaner from a specified period of time, and it understands how those objects have changed".<ref name="extreme" /> ''Engadget'' also noted its unusual capacity for "understanding how telephones and other objects change over time".<ref name="engadget" />

According to ''MIT Technology Review'', one of OpenAI's objectives was to "give language models a better grasp of the everyday concepts that humans use to make sense of things".<ref name="Heaven-2021" />

Wall Street investors have had a positive reception of DALL-E 2, with some firms thinking it could represent a turning point for a future multi-trillion dollar industry. By mid-2019, OpenAI had already received over $1 billion in funding from Microsoft and Khosla Ventures,<ref name="Leswing-2022" /><ref name="Etherington-2019" /><ref name="Fortune" /> and in January 2023, following the launch of DALL-E 2 and ChatGPT, received an additional $10 billion in funding from Microsoft.<ref name="Metz-2023" />

Japan's anime community has had a negative reaction to DALL-E 2 and similar models.<ref name="Rest of World-2022" /><ref name="Roose-2022" /><ref name="Daws-2022" /> Two arguments are typically presented by artists against the software. The first is that AI art is not art because it is not created by a human with intent. "The juxtaposition of AI-generated images with their own work is degrading and undermines the time and skill that goes into their art. AI-driven image generation tools have been heavily criticized by artists because they are trained on human-made art scraped from the web."<ref name="OpenAI-2022b" /> The second is the trouble with copyright law and data text-to-image models are trained on. OpenAI has not released information about what dataset(s) were used to train DALL-E 2, inciting concern from some that the work of artists has been used for training without permission. Copyright laws surrounding these topics are inconclusive at the moment.<ref name="Allyn-2022" />

After integrating DALL-E 3 into Bing Chat and ChatGPT, Microsoft and OpenAI faced criticism for excessive content filtering, with critics saying DALL-E had been "lobotomized."<ref name="Corden-2023" /> The flagging of images generated by prompts such as "man breaks server rack with sledgehammer" was cited as evidence. Over the first days of its launch, filtering was reportedly increased to the point where images generated by some of Bing's own suggested prompts were being blocked.<ref name="Corden-2023" /><ref name="TechRadar" /> ''TechRadar'' argued that leaning too heavily on the side of caution could limit DALL-E's value as a creative tool.<ref name="TechRadar" />

== Open-source implementations == Since OpenAI has not released source code or weights for any of the three models, there have been several attempts to create open-source models offering similar capabilities.<ref name="Mor-2022" /><ref name="Jina-2022" /> Released in 2022 on Hugging Face's Spaces platform, Craiyon (formerly DALL-E Mini until a name change was requested by OpenAI in June 2022) is an AI model based on the original DALL-E that was trained on unfiltered data from the Internet. It attracted substantial media attention in mid-2022, after its release due to its capacity for producing humorous imagery.<ref name="CNETmini" /><ref name="DailyDotmini" /><ref name="Polygonmini" /> Another example of an open source text-to-image model is Stable Diffusion by Stability AI.<ref>{{Citation |title=Stability-AI/stablediffusion |date=2025-04-07 |url=https://github.com/Stability-AI/stablediffusion?tab=readme-ov-file |access-date=2025-04-08 |publisher=Stability AI}}</ref>

== See also == <!---♦♦♦ Please keep the list in alphabetical order by last name ♦♦♦---> * Artificial intelligence art * DeepDream * GPT Image * Imagen * Midjourney * Prompt engineering * Recraft * Runway * Stable Diffusion

==References== {{Reflist|refs=

<ref name="Radford-2018">{{Cite web | url = https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf | title = Improving Language Understanding by Generative Pre-Training | last1 = Radford | first1 = Alec | last2 = Narasimhan | first2 = Karthik | last3 = Salimans | first3 = Tim | last4 = Sutskever | first4 = Ilya | pages = 12 | publisher = OpenAI | date = 11 June 2018 | access-date = 23 January 2021 | archive-date = 26 January 2021 | archive-url = https://web.archive.org/web/20210126024542/https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf | url-status = live}}</ref>

<ref name="Radford-2019">{{cite journal | url = https://cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf | title = Language models are unsupervised multitask learners | last1 = Radford | first1 = Alec | last2 = Wu | first2 = Jeffrey | last3 = Child | first3 = Rewon | last4 = Luan | first4 = David | last5 = Amodei | first5 = Dario | last6 = Sutskever | first6 = Ilua |display-authors=3 | volume = 1 | issue = 8 | date = 14 February 2019 | access-date = 19 December 2020 | quote = | website = cdn.openai.com | archive-date = 6 February 2021 | archive-url = https://web.archive.org/web/20210206183945/https://cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf | url-status = live}}</ref>

<ref name="Brown-2020">{{cite arXiv | last1 = Brown | first1 = Tom B. | last2 = Mann | first2 = Benjamin | last3 = Ryder | first3 = Nick | last4 = Subbiah | first4 = Melanie | last5 = Kaplan | first5 = Jared | last6 = Dhariwal | first6 = Prafulla | last7 = Neelakantan | first7 = Arvind | last8 = Shyam | first8 = Pranav | last9 = Sastry | first9 = Girish | last10 = Askell | first10 = Amanda | last11 = Agarwal | first11 = Sandhini | last12 = Herbert-Voss | first12 = Ariel | last13 = Krueger | first13 = Gretchen | last14 = Henighan | first14 = Tom | last15 = Child | first15 = Rewon | last16 = Ramesh | first16 = Aditya | last17 = Ziegler | first17 = Daniel M. | last18 = Wu | first18 = Jeffrey | last19 = Winter | first19 = Clemens | last20 = Hesse | first20 = Christopher | last21 = Chen | first21 = Mark | last22 = Sigler | first22 = Eric | last23 = Litwin | first23 = Mateusz | last24 = Gray | first24 = Scott | last25 = Chess | first25 = Benjamin | last26 = Clark | first26 = Jack | last27 = Berner | first27 = Christopher | last28 = McCandlish | first28 = Sam | last29 = Radford | first29 = Alec | last30 = Sutskever | first30 = Ilya | last31 = Amodei | first31 = Dario|display-authors=3 | title = Language Models are Few-Shot Learners | eprint = 2005.14165 | date = July 22, 2020 | class = cs.CL}}</ref>

<ref name="dallepaper">{{cite arXiv |last1 = Ramesh |first1 = Aditya |last2 = Pavlov |first2 = Mikhail |last3 = Goh |first3 = Gabriel |last4 = Gray |first4 = Scott |last5 = Voss |first5 = Chelsea |last6 = Radford |first6 = Alec |last7 = Chen |first7 = Mark |last8 = Sutskever |first8 = Ilya|display-authors=3 |eprint = 2102.12092 |title = Zero-Shot Text-to-Image Generation |class = cs.LG |date = 24 February 2021}}</ref>

<ref name="tc">{{cite web | url = https://techcrunch.com/2021/01/05/openais-dall-e-creates-plausible-images-of-literally-anything-you-ask-it-to/ | title = OpenAI's DALL-E creates plausible images of literally anything you ask it to | last = Coldewey | first = Devin | website = | publisher = | date = 5 January 2021 | access-date = 5 January 2021 | quote = | archive-date = 6 January 2021 | archive-url = https://web.archive.org/web/20210106075542/https://techcrunch.com/2021/01/05/openais-dall-e-creates-plausible-images-of-literally-anything-you-ask-it-to/ | url-status = live}}</ref>

<ref name="vb">{{cite web | url = https://venturebeat.com/2021/01/05/openai-debuts-dall-e-for-generating-images-from-text/ | title = OpenAI debuts DALL-E for generating images from text | last = Johnson | first = Khari | website = | publisher = VentureBeat | date = 5 January 2021 | access-date = 5 January 2021 | quote = | archive-date = 5 January 2021 | archive-url = https://web.archive.org/web/20210105221534/https://venturebeat.com/2021/01/05/openai-debuts-dall-e-for-generating-images-from-text/ | url-status = live}}</ref>

<ref name="Heaven-2021">{{cite web | url = https://www.technologyreview.com/2021/01/05/1015754/avocado-armchair-future-ai-openai-deep-learning-nlp-gpt3-computer-vision-common-sense/ | title = This avocado armchair could be the future of AI | last = Heaven | first = Will Douglas | website = | publisher = MIT Technology Review | date = 5 January 2021 | access-date = 5 January 2021 | quote = | archive-date = 5 January 2021 | archive-url = https://web.archive.org/web/20210105193658/https://www.technologyreview.com/2021/01/05/1015754/avocado-armchair-future-ai-openai-deep-learning-nlp-gpt3-computer-vision-common-sense/ | url-status = live}}</ref>

<ref name="boing">{{cite web | url = https://boingboing.net/2021/02/10/this-ai-neural-network-transforms-text-captions-into-art-like-a-jellyfish-pikachu.html | title = This AI neural network transforms text captions into art, like a jellyfish Pikachu | last = Dunn | first = Thom | website = | publisher = BoingBoing | date = 10 February 2021 | access-date = 2 March 2021 | quote = | archive-date = 22 February 2021 | archive-url = https://web.archive.org/web/20210222001459/https://boingboing.net/2021/02/10/this-ai-neural-network-transforms-text-captions-into-art-like-a-jellyfish-pikachu.html | url-status = live}}</ref>

<ref name="nature">{{cite web | url = https://www.nature.com/immersive/d41586-021-00095-y/index.html | title = Tardigrade circus and a tree of life — January's best science images | last = Stove | first = Emma | website = | publisher = Nature | date = 5 February 2021 | access-date = 2 March 2021 | quote = | archive-date = 8 March 2021 | archive-url = https://web.archive.org/web/20210308032636/https://www.nature.com/immersive/d41586-021-00095-y/index.html | url-status = live}}</ref>

<ref name="extreme">{{cite news | url = https://www.extremetech.com/extreme/318881-openais-dall-e-generates-images-from-text-descriptions | title = OpenAI's 'DALL-E' Generates Images From Text Descriptions | last = Whitwam | first = Ryan | newspaper = ExtremeTech | date = 6 January 2021 | access-date = 2 March 2021 | quote = | archive-date = 28 January 2021 | archive-url = https://web.archive.org/web/20210128064428/https://www.extremetech.com/extreme/318881-openais-dall-e-generates-images-from-text-descriptions | url-status = live}}</ref>

<ref name="engadget">{{cite web | url = https://www.engadget.com/dall-e-ai-gpt-make-image-from-any-description-135535140.html | title = OpenAI's DALL-E app generates images from just a description | last = Dent | first = Steve | website = | publisher = Engadget | date = 6 January 2021 | access-date = 2 March 2021 | quote = | archive-date = 27 January 2021 | archive-url = https://web.archive.org/web/20210127225652/https://www.engadget.com/dall-e-ai-gpt-make-image-from-any-description-135535140.html | url-status = live}}</ref>

<ref name="input">{{cite web | url = https://www.inputmag.com/tech/dalle-takes-your-text-turns-it-into-surreal-captivating-art | title = This AI turns text into surreal, suggestion-driven art | last = Kasana | first = Mehreen | website = | publisher = Input | date = 7 January 2021 | access-date = 2 March 2021 | quote = | archive-date = 29 January 2021 | archive-url = https://web.archive.org/web/20210129211643/https://www.inputmag.com/tech/dalle-takes-your-text-turns-it-into-surreal-captivating-art | url-status = live}}</ref>

<ref name="cnbc">{{cite web | url = https://www.cnbc.com/2021/01/08/openai-shows-off-dall-e-image-generator-after-gpt-3.html | title = Why everyone is talking about an image generator released by an Elon Musk-backed A.I. lab | last = Shead | first = Sam | website = | publisher = CNBC | date = 8 January 2021 | access-date = 2 March 2021 | quote = | archive-date = 16 July 2022 | archive-url = https://web.archive.org/web/20220716230354/https://www.cnbc.com/2021/01/08/openai-shows-off-dall-e-image-generator-after-gpt-3.html | url-status = live}}</ref>

<ref name="dale">{{cite web | url = https://thenextweb.com/neural/2021/01/10/heres-how-openais-magical-dall-e-generates-images-from-text-syndication/ | title = Here's how OpenAI's magical DALL-E image generator works | last = Markowitz | first = Dale | website = | publisher = TheNextWeb | date = 10 January 2021 | access-date = 2 March 2021 | quote = | archive-date = 23 February 2021 | archive-url = https://web.archive.org/web/20210223162340/https://thenextweb.com/neural/2021/01/10/heres-how-openais-magical-dall-e-generates-images-from-text-syndication/ | url-status = live}}</ref>

<ref name="nbc">{{cite web | url = https://www.nbcnews.com/tech/innovation/here-s-dall-e-algorithm-learned-draw-anything-you-tell-n1255834 | title = Here's DALL-E: An algorithm learned to draw anything you tell it | last = Ehrenkranz | first = Melanie | website = | publisher = NBC News | date = 27 January 2021 | access-date = 2 March 2021 | quote = | archive-date = 20 February 2021 | archive-url = https://web.archive.org/web/20210220164655/https://www.nbcnews.com/tech/innovation/here-s-dall-e-algorithm-learned-draw-anything-you-tell-n1255834 | url-status = live}}</ref>

<ref name="bbc">{{cite web | url = https://www.bbc.com/news/technology-55559463 | title = AI draws dog-walking baby radish in a tutu | last = Wakefield | first = Jane | website = | publisher = British Broadcasting Corporation | date = 6 January 2021 | access-date = 3 March 2021 | quote = | archive-date = 2 March 2021 | archive-url = https://web.archive.org/web/20210302170623/https://www.bbc.com/news/technology-55559463 | url-status = live}}</ref>

<ref name="cnn">{{cite web | url = https://www.cnn.com/2021/01/08/tech/artificial-intelligence-openai-images-from-text/index.html | title = A radish in a tutu walking a dog? This AI can draw it really well | last = Metz | first = Rachel | website = | publisher = CNN | date = 2 February 2021 | access-date = 2 March 2021 | quote = | archive-date = 16 July 2022 | archive-url = https://web.archive.org/web/20220716171727/https://www.cnn.com/2021/01/08/tech/artificial-intelligence-openai-images-from-text/index.html | url-status = live}}</ref>

<ref name="Knight-2021">{{cite magazine | url = https://www.wired.com/story/ai-go-art-steering-self-driving-car/ | title = This AI Could Go From 'Art' to Steering a Self-Driving Car | last = Knight | first = Will | magazine = Wired | date = 26 January 2021 | access-date = 2 March 2021 | quote = | archive-date = 21 February 2021 | archive-url = https://web.archive.org/web/20210221010223/https://www.wired.com/story/ai-go-art-steering-self-driving-car/ | url-status = live}}</ref>

<ref name="CNETmini">{{cite web | url = https://www.cnet.com/culture/everything-to-know-about-dall-e-mini-the-mind-bending-ai-art-creator/ | title = Everything to Know About Dall-E Mini, the Mind-Bending AI Art Creator | last = Carson | first = Erin | website = CNET | publisher = | date = 14 June 2022 | access-date = 15 June 2022 | quote = | archive-date = 15 June 2022 | archive-url = https://web.archive.org/web/20220615085705/https://www.cnet.com/culture/everything-to-know-about-dall-e-mini-the-mind-bending-ai-art-creator/ | url-status = live}}</ref>

<ref name="DailyDotmini">{{cite web | url = https://www.dailydot.com/unclick/dall-e-mini-memes/ | title = AI program DALL-E mini prompts some truly cursed images | last = Schroeder | first = Audra | website = Daily Dot | publisher = | date = 9 June 2022 | access-date = 15 June 2022 | quote = | archive-date = 10 June 2022 | archive-url = https://web.archive.org/web/20220610212300/https://www.dailydot.com/unclick/dall-e-mini-memes/ | url-status = live}}</ref>

<ref name="Polygonmini">{{cite web | url = https://www.polygon.com/23167596/memes-dall-e-mini-image-generator-ai-explained | title = People are using DALL-E mini to make meme abominations — like pug Pikachu | last = Diaz | first = Ana | website = Polygon | publisher = | date = 15 June 2022 | access-date = 15 June 2022 | quote = | archive-date = 15 June 2022 | archive-url = https://web.archive.org/web/20220615151753/https://www.polygon.com/23167596/memes-dall-e-mini-image-generator-ai-explained | url-status = live}}</ref>

<ref name="Mor-2022">{{cite web | url = https://venturebeat.com/2022/04/16/how-dall-e-2-could-solve-major-computer-vision-challenges/ | title = How DALL-E 2 could solve major computer vision challenges | last = Sahar Mor | first = Stripe | website = VentureBeat | publisher = | date = 16 April 2022 | access-date = 15 June 2022 | quote = | archive-date = 24 May 2022 | archive-url = https://web.archive.org/web/20220524133956/https://venturebeat.com/2022/04/16/how-dall-e-2-could-solve-major-computer-vision-challenges/ | url-status = live}}</ref>

<ref name="Corden-2023">{{cite web | url = https://www.windowscentral.com/software-apps/bing/bing-dall-e-3-image-creation-was-great-for-a-few-days-but-now-microsoft-has-predictably-lobotomized-it | title = Bing Dall-E 3 image creation was great for a few days, but now Microsoft has predictably lobotomized it | last = Corden | first = Jez | website = Windows Central | date = 8 October 2023 | access-date = 11 October 2023 | archive-date = 10 October 2023 | archive-url = https://web.archive.org/web/20231010185641/https://www.windowscentral.com/software-apps/bing/bing-dall-e-3-image-creation-was-great-for-a-few-days-but-now-microsoft-has-predictably-lobotomized-it | url-status = live}}</ref>

<ref name="TechRadar">{{cite web | url = https://www.techradar.com/computing/artificial-intelligence/microsoft-reins-in-bing-ais-image-creator-and-the-results-dont-make-much-sense | title = Microsoft reins in Bing AI's Image Creator – and the results don't make much sense | last = Allan | first = Darren | website = TechRadar | date = 9 October 2023 | access-date = 11 October 2023 | archive-date = 10 October 2023 | archive-url = https://web.archive.org/web/20231010075911/https://www.techradar.com/computing/artificial-intelligence/microsoft-reins-in-bing-ais-image-creator-and-the-results-dont-make-much-sense | url-status = live}}</ref>

<ref name="OpenAI-2022b">{{Cite web | date = 2022-07-20 | title = DALL·E Now Available in Beta | url = https://openai.com/blog/dall-e-now-available-in-beta/ | url-status = live | archive-url = https://web.archive.org/web/20220720162939/https://openai.com/blog/dall-e-now-available-in-beta/ | archive-date = 20 July 2022 | access-date = 2022-07-20 | website = OpenAI | language = en}}</ref>

<ref name="Allyn-2022">{{Cite news | last = Allyn | first = Bobby | date = 2022-07-20 | title = Surreal or too real? Breathtaking AI tool DALL-E takes its images to a bigger stage | language = en | work = NPR | url = https://www.npr.org/2022/07/20/1112331013/dall-e-ai-art-beta-test | url-status = live | access-date = 2022-07-20 | archive-url = https://web.archive.org/web/20220720170255/https://www.npr.org/2022/07/20/1112331013/dall-e-ai-art-beta-test | archive-date = 20 July 2022}}</ref>

<ref name="OpenAI">{{Cite web | title = DALL·E 3 | url = https://openai.com/dall-e-3/ | url-status = live | archive-url = https://web.archive.org/web/20230920225833/https://openai.com/dall-e-3 | archive-date = 20 September 2023 | access-date = 2023-09-21 | website = OpenAI | language = en-US}}</ref>

<ref name="Ramesh-2022">{{Cite arXiv |last1=Ramesh |first1=Aditya |last2=Dhariwal |first2=Prafulla |last3=Nichol |first3=Alex |last4=Chu |first4=Casey |last5=Chen |first5=Mark |date=2022-04-12 |title=Hierarchical Text-Conditional Image Generation with CLIP Latents |class=cs.CV |eprint=2204.06125}}</ref>

<ref name="Marcus-2022">{{cite arXiv |eprint=2204.13807 |class=cs.CV |first1=Gary |last1=Marcus |first2=Ernest |last2=Davis |title=A very preliminary analysis of DALL-E 2 |date=2022-05-02 |last3=Aaronson |first3=Scott}}</ref>

<ref name="OpenAI-2022a">{{Cite web | date = 2022-06-28 | title = DALL·E 2 Pre-Training Mitigations | url = https://openai.com/blog/dall-e-2-pre-training-mitigations/ | access-date = 2022-07-18 | website = OpenAI | language = en | archive-date = 19 July 2022 | archive-url = https://web.archive.org/web/20220719044249/https://openai.com/blog/dall-e-2-pre-training-mitigations/ | url-status = live}}</ref>

<ref name="Taylor">{{cite news | last = Taylor | first = Josh | title = From Trump Nevermind babies to deep fakes: DALL-E and the ethics of AI art | url = https://www.theguardian.com/technology/2022/jun/19/from-trump-nevermind-babies-to-deep-fakes-dall-e-and-the-ethics-of-ai-art | website = The Guardian | date = 18 June 2022 | accessdate = 2 August 2022 | archive-date = 6 July 2022 | archive-url = https://web.archive.org/web/20220706125540/https://www.theguardian.com/technology/2022/jun/19/from-trump-nevermind-babies-to-deep-fakes-dall-e-and-the-ethics-of-ai-art | url-status = live}}</ref>

<ref name="Knight-2022">{{cite magazine | last1 = Knight | first1 = Will | title = When AI Makes Art, Humans Supply the Creative Spark | url = https://www.wired.com/story/when-ai-makes-art/ | magazine = Wired | date = 13 July 2022 | accessdate = 2 August 2022 | archive-date = 2 August 2022 | archive-url = https://web.archive.org/web/20220802162402/https://www.wired.com/story/when-ai-makes-art/ | url-status = live}}</ref>

<ref name="vice">{{cite news | title = DALL-E Is Now Generating Realistic Faces of Fake People | last = Rose | first = Janus | url = https://www.vice.com/en/article/dall-e-is-now-generating-realistic-faces-of-fake-people/ | date = 24 June 2022 | work = Vice | access-date = 2 August 2022 | archive-date = 30 July 2022 | archive-url = https://web.archive.org/web/20220730133250/https://www.vice.com/en/article/g5vbx9/dall-e-is-now-generating-realistic-faces-of-fake-people | url-status = live}}</ref>

<ref name="docs">{{cite web | title = DALL·E 2 Preview – Risks and Limitations | author = OpenAI | website = GitHub | url = https://github.com/openai/dalle-2-preview/blob/main/system-card.md | date = 19 June 2022 | accessdate = 2 August 2022 | archive-date = 2 August 2022 | archive-url = https://web.archive.org/web/20220802162403/https://github.com/openai/dalle-2-preview/blob/main/system-card.md | url-status = live}}</ref>

<ref name="David-2023">{{Cite web | last = David | first = Emilia | date = 2023-09-20 | title = OpenAI releases third version of DALL-E | url = https://www.theverge.com/2023/9/20/23881241/openai-dalle-third-version-generative-ai | access-date = 2023-09-21 | website = The Verge | language = en-US | archive-date = 20 September 2023 | archive-url = https://web.archive.org/web/20230920192429/https://www.theverge.com/2023/9/20/23881241/openai-dalle-third-version-generative-ai | url-status = live}}</ref>

<ref name="platform.openai.com">{{Cite web | title = OpenAI Platform | url = https://platform.openai.com/ | access-date = 2023-11-10 | website = platform.openai.com | language = en | archive-date = 20 March 2023 | archive-url = https://web.archive.org/web/20230320023933/https://platform.openai.com/ | url-status = live}}</ref>

<ref name="Niles-2023">{{Cite web | last = Niles | first = Raymond | date = 2023-11-10 | orig-date = Updated this week | title = DALL-E 3 API | url = https://help.openai.com/en/articles/8555480-dall-e-3-api | access-date = 2023-11-10 | website = OpenAI help Center | language = en | archive-date = 10 November 2023 | archive-url = https://web.archive.org/web/20231110182305/https://help.openai.com/en/articles/8555480-dall-e-3-api | url-status = live}}</ref>

<ref name="Mehdi-2023">{{Cite web | last = Mehdi | first = Yusuf | date = 2023-09-21 | title = Announcing Microsoft Copilot, your everyday AI companion | url = https://blogs.microsoft.com/blog/2023/09/21/announcing-microsoft-copilot-your-everyday-ai-companion/ | access-date = 2023-09-21 | website = The Official Microsoft Blog | language = en-US | archive-date = 21 September 2023 | archive-url = https://web.archive.org/web/20230921150139/https://blogs.microsoft.com/blog/2023/09/21/announcing-microsoft-copilot-your-everyday-ai-companion/ | url-status = live}}</ref>

<ref name="OpenAI-2">{{Cite web | title = DALL·E 2 | url = https://openai.com/dall-e-2/ | url-status = live | archive-url = https://web.archive.org/web/20220406141035/https://openai.com/dall-e-2/ | archive-date = 6 April 2022 | access-date = 2022-07-06 | website = OpenAI | language = en-US}}</ref>

<ref name="labs.openai.com">{{Cite web | title = DALL·E Waitlist | url = https://labs.openai.com/ | url-status = live | archive-url = https://web.archive.org/web/20220704184756/https://labs.openai.com/ | archive-date = 4 July 2022 | access-date = 2022-07-06 | website = labs.openai.com | language = en}}</ref>

<ref name="Guardian-2022">{{Cite web | date = 2022-06-18 | title = From Trump Nevermind babies to deep fakes: DALL-E and the ethics of AI art | url = https://www.theguardian.com/technology/2022/jun/19/from-trump-nevermind-babies-to-deep-fakes-dall-e-and-the-ethics-of-ai-art | url-status = live | archive-url = https://web.archive.org/web/20220706125540/https://www.theguardian.com/technology/2022/jun/19/from-trump-nevermind-babies-to-deep-fakes-dall-e-and-the-ethics-of-ai-art | archive-date = 6 July 2022 | access-date = 2022-07-06 | website = the Guardian | language = en}}</ref>

<ref name="OpenAI-2022c">{{Cite web | date = 2022-09-28 | title = DALL·E Now Available Without Waitlist | url = https://openai.com/blog/dall-e-now-available-without-waitlist/ | url-status = live | archive-url = https://web.archive.org/web/20221004093145/https://openai.com/blog/dall-e-now-available-without-waitlist/ | archive-date = 4 October 2022 | access-date = 2022-10-05 | website = OpenAI | language = en}}</ref>

<ref name="OpenAI-2022d">{{Cite web | date = 2022-11-03 | title = DALL·E API Now Available in Public Beta | url = https://openai.com/blog/dall-e-api-now-available-in-public-beta | url-status = live | archive-url = https://web.archive.org/web/20221119222937/https://openai.com/blog/dall-e-api-now-available-in-public-beta/ | archive-date = 19 November 2022 | access-date = 2022-11-19 | website = OpenAI | language = en}}</ref>

<ref name="Wiggers-2022">{{Cite news | last = Wiggers | first = Kyle | date = 2022-11-03 | title = Now anyone can build apps that use DALL-E 2 to generate images | work = TechCrunch | url = https://techcrunch.com/2022/11/03/now-anyone-can-build-apps-that-use-dall-e-2-to-generate-images | url-status = live | access-date = 2022-11-19 | archive-url = https://web.archive.org/web/20221119154222/https://techcrunch.com/2022/11/03/now-anyone-can-build-apps-that-use-dall-e-2-to-generate-images/ | archive-date = 19 November 2022}}</ref>

<ref name="GPT-2023">{{cite web | url = https://www.makeuseof.com/gpt-models-explained-and-compared/ | title = GPT-1 to GPT-4: Each of OpenAI's GPT Models Explained and Compared | date = 11 April 2023 | access-date = 29 April 2023 | archive-date = 15 April 2023 | archive-url = https://web.archive.org/web/20230415175013/https://www.makeuseof.com/gpt-models-explained-and-compared/ | url-status = live}}</ref>

<ref name="OpenAI-2021">{{Cite web | date = 2021-01-05 | title = DALL·E: Creating Images from Text | url = https://openai.com/blog/dall-e/ | access-date = 2022-08-13 | website = OpenAI | language = en | archive-date = 27 March 2021 | archive-url = https://web.archive.org/web/20210327133043/https://openai.com/blog/dall-e/ | url-status = live}}</ref>

<ref name="Edwards-2023">{{Cite web | last = Edwards | first = Benj | date = 2023-09-20 | title = OpenAI's new AI image generator pushes the limits in detail and prompt fidelity | url = https://arstechnica.com/information-technology/2023/09/openai-announces-dall-e-3-a-next-gen-ai-image-generator-based-on-chatgpt/ | access-date = 2023-09-21 | website = Ars Technica | language = en-us | archive-date = 21 September 2023 | archive-url = https://web.archive.org/web/20230921130853/https://arstechnica.com/information-technology/2023/09/openai-announces-dall-e-3-a-next-gen-ai-image-generator-based-on-chatgpt/ | url-status = live}}</ref>

<ref name="Coldewey-2022">{{Cite web | last = Coldewey | first = Devin | date = 2022-04-06 | title = New OpenAI tool draws anything, bigger and better than ever | url = https://techcrunch.com/2022/04/06/openais-new-dall-e-model-draws-anything-but-bigger-better-and-faster-than-before/ | access-date = 2022-11-26 | website = TechCrunch | language = en-US | archive-date = 6 May 2023 | archive-url = https://web.archive.org/web/20230506073242/https://techcrunch.com/2022/04/06/openais-new-dall-e-model-draws-anything-but-bigger-better-and-faster-than-before/ | url-status = live}}</ref>

<ref name="OpenAI-2022">{{Cite web | date = 2022-08-31 | title = DALL·E: Introducing Outpainting | url = https://openai.com/blog/dall-e-introducing-outpainting/ | access-date = 2022-11-26 | website = OpenAI | language = en | archive-date = 26 November 2022 | archive-url = https://web.archive.org/web/20221126173545/https://openai.com/blog/dall-e-introducing-outpainting/ | url-status = live}}</ref>

<ref name="Saharia-2022">{{cite arXiv |eprint=2205.11487 |class=cs.CV |first1=Chitwan |last1=Saharia |first2=William |last2=Chan |title=Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding |date=2022-05-23 |last3=Saxena |first3=Saurabh |last4=Li |first4=Lala |last5=Whang |first5=Jay |last6=Denton |first6=Emily |last7=Ghasemipour |first7=Seyed Kamyar Seyed |last8=Ayan |first8=Burcu Karagol |last9=Mahdavi |first9=S. Sara |last10=Lopes |first10=Rapha Gontijo |last11=Salimans |first11=Tim |display-authors=3}}</ref>

<ref name="Marcus-2022a">{{Cite web | last = Marcus | first = Gary | date = 2022-05-28 | title = Horse rides astronaut | url = https://garymarcus.substack.com/p/horse-rides-astronaut | access-date = 2022-06-18 | website = The Road to AI We Can Trust | archive-date = 19 June 2022 | archive-url = https://web.archive.org/web/20220619135711/https://garymarcus.substack.com/p/horse-rides-astronaut | url-status = live}}</ref>

<ref name="Strickland-2022">{{Cite web | last = Strickland | first = Eliza | date = 2022-07-14 | title = DALL-E 2's Failures Are the Most Interesting Thing About It | url = https://spectrum.ieee.org/openai-dall-e-2 | access-date = 2022-08-16 | website = IEEE Spectrum | language = en | archive-date = 15 July 2022 | archive-url = https://web.archive.org/web/20220715204154/https://spectrum.ieee.org/openai-dall-e-2 | url-status = live}}</ref>

<ref name="Vincent-2022">{{cite web | url = https://www.theverge.com/2022/9/28/23376328/ai-art-image-generator-dall-e-access-waitlist-scrapped | author = James Vincent | date = September 29, 2022 | title = OpenAI's image generator DALL-E is available for anyone to use immediately | website = The Verge | access-date = 29 September 2022 | archive-date = 29 September 2022 | archive-url = https://web.archive.org/web/20220929061004/https://www.theverge.com/2022/9/28/23376328/ai-art-image-generator-dall-e-access-waitlist-scrapped | url-status = live}}</ref>

<ref name="Lane-2022">{{cite magazine | last1 = Lane | first1 = Laura | title = DALL-E, Make Me Another Picasso, Please | url = https://www.newyorker.com/magazine/2022/07/11/dall-e-make-me-another-picasso-please | magazine = The New Yorker | date = 1 July 2022 | accessdate = 2 August 2022 | archive-date = 2 August 2022 | archive-url = https://web.archive.org/web/20220802162403/https://www.newyorker.com/magazine/2022/07/11/dall-e-make-me-another-picasso-please | url-status = live}}</ref>

<ref name="Goldman-2022">{{cite web | title = OpenAI: Will DALL-E 2 kill creative careers? | last = Goldman | first = Sharon | date = 26 July 2022 | url = https://venturebeat.com/business/openai-will-dall-e-2-kill-creative-careers/ | access-date = 16 August 2022 | archive-date = 15 August 2022 | archive-url = https://web.archive.org/web/20220815014607/https://venturebeat.com/business/openai-will-dall-e-2-kill-creative-careers/ | url-status = live}}</ref>

<ref name="Blain-2022">{{cite web | title = DALL-E 2: A dream tool and an existential threat to visual artists | last = Blain | first = Loz | date = 29 July 2022 | url = https://newatlas.com/computers/dall-e-2-ai-art/ | access-date = 16 August 2022 | archive-date = 17 August 2022 | archive-url = https://web.archive.org/web/20220817064216/https://newatlas.com/computers/dall-e-2-ai-art/ | url-status = live}}</ref>

<ref name="Leswing-2022">{{Cite web | last = Leswing | first = Kif | title = Why Silicon Valley is so excited about awkward drawings done by artificial intelligence | url = https://www.cnbc.com/2022/10/08/generative-ai-silicon-valleys-next-trillion-dollar-companies.html | access-date = 2022-12-01 | website = CNBC | date = 8 October 2022 | language = en | archive-date = 29 July 2023 | archive-url = https://web.archive.org/web/20230729005158/https://www.cnbc.com/2022/10/08/generative-ai-silicon-valleys-next-trillion-dollar-companies.html | url-status = live}}</ref>

<ref name="Etherington-2019">{{Cite web | last = Etherington | first = Darrell | date = 2019-07-22 | title = Microsoft invests $1 billion in OpenAI in new multiyear partnership | url = https://techcrunch.com/2019/07/22/microsoft-invests-1-billion-in-openai-in-new-multiyear-partnership/ | access-date = 2023-09-21 | website = TechCrunch | language = en-US | archive-date = 22 July 2019 | archive-url = https://web.archive.org/web/20190722173752/https://techcrunch.com/2019/07/22/microsoft-invests-1-billion-in-openai-in-new-multiyear-partnership/ | url-status = live}}</ref>

<ref name="Fortune">{{Cite web | title = OpenAI's first VC backer weighs in on generative A.I. | url = https://fortune.com/2023/02/02/openais-first-vc-backer-khosla-ventures-weighs-in-on-the-future-of-generative-a-i/ | access-date = 2023-09-21 | website = Fortune | language = en | archive-date = 23 October 2023 | archive-url = https://web.archive.org/web/20231023211810/https://fortune.com/2023/02/02/openais-first-vc-backer-khosla-ventures-weighs-in-on-the-future-of-generative-a-i/ | url-status = live}}</ref>

<ref name="Metz-2023">{{Cite news | last1 = Metz | first1 = Cade | last2 = Weise | first2 = Karen | date = 2023-01-23 | title = Microsoft to Invest $10 Billion in OpenAI, the Creator of ChatGPT | language = en-US | work = The New York Times | url = https://www.nytimes.com/2023/01/23/business/microsoft-chatgpt-artificial-intelligence.html | access-date = 2023-09-21 | issn = 0362-4331 | archive-date = 21 September 2023 | archive-url = https://web.archive.org/web/20230921171844/https://www.nytimes.com/2023/01/23/business/microsoft-chatgpt-artificial-intelligence.html | url-status = live}}</ref>

<ref name="Rest of World-2022">{{Cite web | date = 2022-10-27 | title = AI-generated art sparks furious backlash from Japan's anime community | url = https://restofworld.org/2022/ai-backlash-anime-artists/ | access-date = 2023-01-03 | website = Rest of World | language = en-US | archive-date = 31 December 2022 | archive-url = https://web.archive.org/web/20221231155114/https://restofworld.org/2022/ai-backlash-anime-artists/ | url-status = live}}</ref>

<ref name="Roose-2022">{{Cite news | last = Roose | first = Kevin | date = 2022-09-02 | title = An A.I.-Generated Picture Won an Art Prize. Artists Aren't Happy. | language = en-US | work = The New York Times | url = https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html | access-date = 2023-01-03 | issn = 0362-4331 | archive-date = 31 May 2023 | archive-url = https://web.archive.org/web/20230531203858/https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html | url-status = live}}</ref>

<ref name="Daws-2022">{{Cite web | last = Daws | first = Ryan | date = 2022-12-15 | title = ArtStation backlash increases following AI art protest response | url = https://www.artificialintelligence-news.com/2022/12/15/artstation-backlash-increases-ai-art-protest-response/ | access-date = 2023-01-03 | website = AI News | language = en-GB | archive-date = 3 January 2023 | archive-url = https://web.archive.org/web/20230103233938/https://www.artificialintelligence-news.com/2022/12/15/artstation-backlash-increases-ai-art-protest-response/ | url-status = live}}</ref>

<ref name="Jina-2022">{{cite web | title = jina-ai/dalle-flow | date = 2022-06-17 | url = https://github.com/jina-ai/dalle-flow | publisher = Jina AI | access-date = 2022-06-17 | archive-date = 17 June 2022 | archive-url = https://web.archive.org/web/20220617090248/https://github.com/jina-ai/dalle-flow | url-status = live}}</ref> }}

== External links == {{Commons category}} * {{cite arXiv |last1=Ramesh |first1=Aditya |title=Zero-Shot Text-to-Image Generation |date=2021-02-26 |eprint=2102.12092 |last2=Pavlov |first2=Mikhail |last3=Goh |first3=Gabriel |last4=Gray |first4=Scott |last5=Voss |first5=Chelsea |last6=Radford |first6=Alec |last7=Chen |first7=Mark |last8=Sutskever |first8=Ilya|class=cs.CV }}. The original report on DALL-E. * [https://cdn.openai.com/papers/DALL_E_3_System_Card.pdf DALL-E 3 System Card] * [https://cdn.openai.com/papers/dall-e-3.pdf DALL-E 3] paper by OpenAI * [https://openai.com/dall-e-2/ DALL-E 2 website] * [https://www.craiyon.com/ Craiyon website]

{{OpenAI navbox}} {{Artificial intelligence navbox}} {{Generative AI}} {{Authority control}}

Category:Text-to-image generation Category:Deep learning software applications Category:Unsupervised learning Category:Generative pre-trained transformers Category:2021 software Category:ChatGPT Category:2021 in artificial intelligence