Generative AI’s breakout year: Digiday’s definitive 2023 timeline

Rather than writing a lengthy intro about 2023, we thought it made sense to have bot — in this case Google Bard — distill our timeline as a sort of TL/DR. (We also had the bot write a poem about our timeline in the style of the Bard of Avalon, William Shakespeare.) Here’s what it wrote when asked for a paragraph summary of our timeline:

“The year 2023 was a landmark period in the field of artificial intelligence, characterized by a surge in generative AI advancements and adoption across various sectors, led by significant events such as Microsoft’s investment in OpenAI and the launch of GPT-4. This rapid growth, however, was paralleled by rising ethical, legal, and regulatory challenges, highlighted by copyright infringement lawsuits against AI companies and increased government involvement, including the U.S. Senate’s AI hearings and President Biden’s executive order. These developments collectively signaled a shift in the AI landscape, where technological innovation is increasingly intertwined with calls for responsible development and oversight.”

January:

Microsoft further fueled the generative AI boom when it announced an extended partnership with OpenAI after weeks of reports that it had invested $10 billion into the then-small research startup. (Days before its big announcement, MSFT announced plans to lay off 10,000 employees.)

Other January news

• Shutterstock added new generative AI features into its platform.
• During CES 2023, Nvidia debuted new generative AI technologies for its Omniverse platform to help create virtual worlds faster.
• Getty Images filed a lawsuit against Stability AI, alleging the text-to-image startup illegally copied and processed millions of images without permission. Another lawsuit against Stability AI and Midjourney was filed the same month by a group of artists, who also alleged copyright infringement.

February:

Momentum for generative AI began building beyond just tech giants. As Meta and Snapchat rolled out new tools, agencies — including Horizon, DDB and Havas —  unveiled new generative AI capabilities for clients. Meanwhile, a wave of new AI-based marketing startups also looked for ways to compete. At the same time, more experts warned of the potentially major risks generative AI posed for misinformation, cybersecurity, data privacy and fraud.

Others February news

• Bard, Google’s own chatbot, was introduced along with a number of other AI-related updates from the search giant. 
• Typeface, a generative AI content creation startup, emerged from stealth mode with a $65 million fundraising round.
• Microsoft previewed a newly updated Bing while also rolling out new AI tools for Skype, Edge and Github.

March:

March was another big month for OpenAI. It debuted the long awaited GPT-4, rolled out new enterprise-grade tools and opened the AI floodgates with a new API with early partners like Snap, Instacart, and Shopify. A few weeks later, it introduced ChatGPT plugins starting with a wide range of brands including Klarna, Expedia, Kayak and OpenTable. In some ways, the APIs and plugins seemed to follow a familiar playbook similar to Google and Facebook’s APIs during their early days. However, ChatGPT also was bogged down by an outage along with some users having payment data leaked by what OpenAI described as a bug.

Other March news:

• March was also a big month for AI-enabled design. Adobe officially debuted its Firefly generative AI platform, while Canva rolled out its own suite of new generative AI tools. Meanwhile, Nvidia announced new partnerships with several companies including Adobe, Getty and Shutterstock.
• Google added new generative AI tools for text and images and other features for Gmail and Google Docs.
• Salesforce debuted Einstein GPT, a new AI model built specifically for CRM clients. The company also announced a new $250 million Generative AI fund through Salesforce Ventures.
• In the gaming world, Unity opened a new generative AI marketplace for gaming and Roblox debuted its own new AI tools.

April:

After the March debut of Midjourney v5, the next version of the popular text-to-image AI platform, photorealistic images quickly impressed many users and viewers. But the heightened curiosity surrounding some viral AI-made images — including one portraying Pope Francis wearing a Balenciaga jacket — also prompted new worries about more misinformation coming from AI-generated content.

Other April news

• Snap began offering access to its “My AI” chatbot to all Snapchat users globally, expanding beyond paid subscribers who gained access in February. (Since then, Snap has added a number of other new AI text and image features for users.)
• Reports of Google planning an overhaul of search with more AI tools had marketers wondering about the potential impact for search marketing.
• Alibaba announced a new AI model to compete with U.S. chatbots like ChatGPT and Chinese giants like Baidu. The chatbot’s name, Tongyi Qianwen, means “truth from a thousand questions.”

May:

Discussions about AI arrive in Congress as the U.S. Senate held its first AI hearing, which was followed up by other hearings and forums in following months. The first hearing was also a first for OpenAI cofounder and CEO Sam Altman, who made his debut as a witness alongside IBM chief privacy and trust officer Christina Montgomery and NYU professor Gary Marcus.

Other May news

• Microsoft teased out a new chat ads API for publishers, which was followed up months later with news of Snap and Axel Springer as early partners.
• Companies and organizations expanded efforts to prevent AI risks as TruePic partnered with the World Economic Forum and Shutterstock announced a new “AI For Good” initiative with the United Nations.
• The neighborhood-focused social network Nextdoor debuted a new AI assistant to help users give business recommendations and write posts.
• Snowflake acquired Neeva — a search startup without advertising — to boost the cloud company’s generative AI capabilities.
• During strikes organized by WGA and SAG-AFTRA — which included pickets of various Upfronts — generative AI was a key concern for writers and actors. 

June:

Major companies committed billions of dollars more to accelerate their efforts in generative AI. Following PwC’s commitment to spend $1 billion over the next three years, Accenture announced plans to spent $3 billion over the same period. Meanwhile, Salesforce said it would raise its investment commitment for generative AI startups from $250 million to $500 million. (Earlier in the year, Salesforce’s Generative AI Fund had already invested in a number of startups including Anthropic, the search company You.com and the enterprise-focused Cohere.)

Other June news

• Amid growing concerns about AI’s impact on intellectual property, two new lawsuits were filed against OpenAI alleging IP infringement. (One filed by Clarkson Law Firm was then later dismissed in September.)
• Giants like Shutterstock and Adobe began offering IP indemnity to clients if they face legal challenges from using their AI products.
• Inflection AI — another AI startup cofounded by LinkedIn Cofounder Reid Hoffman and DeepMind Cofounder Mustafa Suleyman — raised $1.3 billion from Nvidia and others as part of a complain to compete with OpenAI and others.
• U.S. Senator Chuck Schumer introduced new legislation called the “SAFE Innovation Framework” to address safe AI development. He also announced plans to host several meetings with AI leaders in a series called the “AI Insight Forum” which included top tech executives and other experts. (Some criticized the series for not having sufficient diversity of experts.)

July:

Meta and Microsoft made a mid-summer splash with Llama 2, a new open-source large language model that the companies said would be free for research and most commercial use.

Other July news

• Days after meeting with the White House to discuss safe AI development, several major AI players — including Microsoft, Anthropic, Google and OpenAI — announced a new organization called the Frontier Model Forum to self-regulate AI development. 
• Days before the Senate Judiciary Committee held another hearing to discuss copyright issues regarding AI, lawmakers introduced a bipartisan AI bill called the U.S. AI Act. 
• The Federal Trade Commission also opened an investigation into OpenAI to see how it gathers and protects data.

August:

AI and other new tech showed up plenty during the 2023 FIFA Women’s World Cup. FIFA used AI to moderate content and even create a new game with AI-generated players. Meanwhile, the French telecom Orange used a deepfake for an ad about sexism in soccer and Cadbury created a new tool for fans to create posters.

Other August news

• IBM and Salesforce announced a partnership to tap into both companies’ technologies by integrating Salesforce’s Einstein with IBM’s WatsonX platform for tasks such as data mapping and new integrations across Salesforce’s Sales Cloud, Marketing Cloud and Slack.
• More companies began building their own AI assistants powered by large language models. For example, Publicis-owned Profitero debuted a GPT-powered chatbot called “Ask Profitero” for e-commerce data.
• Rapid AI rollouts led to a growing procurement headache as marketers grappled with how to sift through all the noise. 

September:

Weeks after Meta announced a number of new generative AI features for its various apps — including plans for human-like chatbots with the likenesses of real celebrities — Microsoft announced new AI “copilots” and its new AI-powered Surface laptops. Meantime, Google connected its Bard chatbot to more apps including YouTube and Gmail. Amazon also announced several big AI updates for its hardware and software products including a new LLM for Alexa and various smart devices that will be powered by it. 

Other September news

• Getty Images debuted its first AI image platform just a few weeks after Bria — an Israeli AI image startup — unveiled a new AI model trained on licensed images from Getty Images and providers. (Getty became an investor in Bria last year.)
• British Parliament hosted an AI hearing with a number of expert witnesses while the the UK’s Competition and Marketers Authority issued a new report examining how AI foundation models might impact consumers and competition.
• OpenAI debuted DALL-E 3, the next-generation version of its popular AI image generator, and announced ChatGPT could finally browse the internet. It also announced new multimodal capabilities for ChatGPT to help it “hear,” speak” and “see.”
• Spotify announced a new AI voice translation pilot to help podcasters reach audiences in more languages.
• Anthropic introduced Claude Pro, a subscription version of its chatbot, which will give more features to users who pay $20 a month.
• Eight more tech giants made voluntary commitments to join the White House’s pledge related to the responsible AI development.

October:

In a milestone for AI policy in the U.S., President Joe Biden signed a new executive order with the goal of avoiding AI risks while also fostering innovation. Describing AI as the “most consequential technology of our time,” Biden said the there is “no greater change that I can think of in my life that AI presents as a potential [for] exploring the universe, fighting climate change, ending cancer as we know it, and so much more.”

Other October news

• AI-related risks for social media were front and center on a number of fronts. As deepfakes ads of Tom Hanks, Gayle King and MrBeast showcased the danger of misinformation, UK’s Information Commissioner’s Office issued a preliminary enforcement notice to Snap saying the platform hadn’t assessed the risks posed to kids who use its “My AI” chatbot.
• Universal Music Group and other music publishers filed a new lawsuit against Anthropic claiming its Claude chatbot violated IP laws by training its LLM with copyrighted music lyrics.
• Klarna debuted a new AI-powered “shopping lens” to aid visual search, which followed several other AI-related updates by the Swedish fintech firm.
• Qualcomm hosted its annual Snapdragon Summit, where it touted new on-device AI capabilities for PCs and mobile phones that use Qualcomm’s chips.
• A new AI model called Latimer debuted with the goal of mitigating AI’s problem with racial bias. Named after the 19th century Black inventor Lewis Latimer, the startup trained its LLM on a more diverse set of cultural and historical data. It’s also working with students and professors at Historically Black Colleges and Universities.

November:

OpenAI’s board suddenly ousted founder and CEO Sam Altman. A week later, OpenAI’s board agreed to rehire him along with president Greg Brockman and staving off a potential mass exodus of hundreds of OpenAI employees who threatened to quit if Altman wasn’t reinstated. Although he was soon rehired, it prompted plenty of new questions about what the startup’s suddenly mercurial state might mean for the future of generative AI. 

Other November news

• A year ahead of the 2024 U.S. elections, Meta said it would require political advertisers to disclose when images, videos or audio are altered with AI. The updated policies — announced Nov. 8 and updated weeks later — follow similar policy changes Google announced in September. 
• X, formerly known as Twitter, introduced its own AI model called Grok, which was trained on user content and aims to compete with ChatGPT and other chatbots while offering answers based on real-time information.
• As part of the deal to bring back Altman, OpenAI added new board members including Bret Taylor — the former co-CEO of Salesforce who was also Twitter’s board chair when it was acquired by Elon Musk — and former U.S. Treasury secretary Larry Summers. (OpenAI also faced criticism for not adding any women or people of color even after the two women who were on the board left.) 
• Amazon introduced a new AI chatbot named Q, which was built to help enterprise businesses with problem-solving and other tasks.

December:

Google announced its next major AI model called Gemini, which will power a range of Google products and services. (The big debut also capped off a year full of new names of various AI companies, products and platforms that have entered the cultural lexicon.) Google also faced criticism for seemingly faking part of a demo during Gemini’s big reveal.

Other December news

• Musk’s AI startup, xAI, disclosed plans to raise $1 billion in new funding. In paperwork filed Dec. 5 with the U.S. Securities and Exchange Commission, the company said it had already raised $135 million.
• British regulators launched a new inquiry into how Microsoft’s partnership with OpenAI impacts competition in the broader market.
• Amazon debuted new training tools.
• The European Union reached an agreement on the proposed AI Act, which will include a range of new AI-related regulations that will go into effect in 2025.