The hype around ChatGPT at the end of 2022 became a catalyst for a race of providers around Generative AI: Everyone was in a hurry to launch their AI tools: Google published Bard, Microsoft countered with Bing Chat (now Copilot). The development momentum was enormous: OpenAI churned out new releases, and competitors from Google to Llama managed an impressive race to catch up with the market leader ChatGPT.

Equally noteworthy: Many of the leading AI models, such as OpenAI‘s GPT series, started the year as text-only models and ended it as multimodal models capable of processing and producing images, audio, and in some cases, even video. As models become more powerful, multimodality is expected to become more intuitive and commonplace in 2024.

2023 will also go down in the history books as the year in which the first promising results were achieved in AI-based video generation. Below is an overview on the timeline:

Source: Justine Moore (@venturetwins on X)

The AI hype has also raised fears about the rationalization of jobs. This concern was particularly evident in the strike of the Writers Guild of America, which paralyzed film production in the USA from May to September. As a result, the writers’ association won important concessions from the film studios, including assurances that AI would not be used to write or train their work. While that’s a big win for writers, it’s probably not the last conflict, given how fast AI is evolving and how many creative fields it will affect.

Politicians have already responded to fears about AI – at the end of 2023, three key measures in particular are important: First, President Biden issued an executive order to protect Americans and control risks. Second, shortly thereafter, the UK hosted an AI Security Summit where all countries signed a declaration highlighting the risks posed by powerful AI, emphasizing the responsibility of AI developers for the security of their systems, and pledging international cooperation to mitigate these risks. Thirdly (and certainly the policy measure with the biggest impact), the EU AI Act, the world’s first comprehensive legal framework for AI. The act, adopted by the European Parliament in June, is currently still being negotiated between the European Parliament, the European Council and the European Commission in a procedure known as a trilogue before it can enter into force.

2023: AI development milestones in 2023 – month-by-month

January

  • Microsoft is making waves with a huge $10 billion investment in OpenAI
  • February

  • Meta launches Llama 2, the open-source competitor to ChatGPT
  • OpenAI announces ChatGPT Plus, a paid version of their chatbot
  • Note: Open AI generated 1.6bn USD in annual revenue in 2023 – a huge jump from last year’s 28 mUSD. The huge increase is primarily due to the paid version of ChatGPT, which was launched in early 2023.

    March

  • OpenAI announces the high-performance GPT-4 model, which is still considered the gold standard
  • Midjourney Releases V5, Bringing AI-Powered Image Generation One Step Closer to Reality
  • Google launches Bard, its counterpart to ChatGPT
  • Microsoft launches Copilot for Microsoft 365
  • Think of Copilot as a sophisticated digital assistant that’s not just for chat, but integrates with a range of Microsoft apps like Word, Teams, and even the operating system. It’s pretty handy – it can output images, summarize meetings, retrieve information, and even sync things across devices – all with the goal of making users’ lives easier by automating the annoying stuff.

    April

  • Elon Musk, Steve Wozniak, and more than 1,100 other people sign a petition calling for the training of high-capacity AI models that are more powerful than GPT-4 to be stopped
  • May

  • Samsung and other companies prohibit their employees from using Gen AI tools due to security concerns, compare the article Samsung bans its employees from using ChatGPT
  • OpenAI releases ChatGPT plugins and web browsing features
  • Nvidia’s stock price soars 27% overnight, reaching a market cap of nearly 1 tn USD, thanks to booming demand for GPUs needed to train AI models>
  • June

  • Adobe announces Firefly, a powerful image generation tool
  • Accenture announces a staggering 3 bn USD investment in AI
  • July

  • ChatGPT Releases Code Interpreter and Data Analysis Capabilities
  • Stack Overflow’s traffic drops by half due to Gen AI encoding tools, see also article AI-based encoding wizards: How good are they really?
  • August

  • Salesforce reportedly leads OpenAI competitor Hugging Face’s funding with a valuation of over 4 bn USD
  • OpenAI releases ChatGPT Enterprise for enterprise
  • September

  • OpenAI releases Dall-E 3, a powerful new image generation tool
  • OpenAI also introduces multimodal capabilities for ChatGPT
  • Meta releases AI chatbots for its messaging platforms, smart glasses, AI-powered image editing, and more
  • October

  • Google launches new Pixel phones with Gen AI capabilities
  • 90s nostalgia app Epik uses AI to top the app charts
  • Chinese company Baidu launches a ChatGPT competitor
  • November

  • Elon Musk launches Grok, his ChatGPT competitor
  • OpenAI kündigt Custom GPTs und GPT-4 Turbo an
  • ex-Apple employees unveil the Humane Ai Pin, an AI wearable device
  • Nvidia announces H200 chips that will power the next generation of AI
  • Sam Altman drama: Sam Altman, the face of OpenAI and the entire AI industry, was unexpectedly fired from OpenAI in November. The reason? – Altman wasn’t always transparent in his… Explanations? Shops? Neither Ilya nor the board of directors really let us know – with the company. But the drama didn’t stop there. The company’s employees were up in arms and were ready to leave the company if Altman was not brought back. Meanwhile, OpenAI investor Microsoft stepped in and offered jobs to Altman and any OpenAI employees who were willing to leave the company. For a moment, it looked like OpenAI was just falling apart. But then, suddenly, Altman was back at it.

    December

  • Pika Labs ushers in a new era of AI video production with Pika 1.0
  • Google launches Gemini, its AI model that can supposedly beat GPT-4
  • Midjourney takes AI imagery to the next level with its V6 update
  • Author

    The author is a manager in the software industry with international expertise: Authorized officer at one of the large consulting firms - Responsible for setting up an IT development center at the Bangalore offshore location - Director M&A at a software company in Berlin.