AI: Every Crazy Thing That Happened in 2023

1. AI advancements in 2023 were numerous and significant.

🥇95 00:00

Throughout the year, there were hundreds, if not thousands, of major AI announcements and advancements.

• The year saw major investments, product launches, and breakthroughs in various AI domains.
• The pace of AI development and innovation is accelerating rapidly.

2. Microsoft's $10 billion investment in OpenAI was a major move.

🥈88 01:14

Microsoft's investment in OpenAI allowed them to incorporate OpenAI's technology into their products.

• This strategic move strengthened Microsoft's position in the AI market.
• The partnership between Microsoft and OpenAI has resulted in significant advancements in AI capabilities.

3. The launch of 11 Labs beta improved text-to-speech technology.

🥈82 02:27

11 Labs beta introduced high-quality text-to-speech capabilities.

• Prior to 11 Labs, text-to-speech sounded robotic and artificial.
• The improved text-to-speech technology had a significant impact on the AI art world.

4. Instruct Picks to Pi revolutionized AI image editing.

🥈87 04:00

Instruct Picks to Pi allowed for advanced image editing using AI.

• This release enabled users to generate and modify images with unprecedented control.
• AI-generated images became more realistic and customizable.

5. Google's release of Google Bard aimed to compete with OpenAI.

🥉79 07:51

Google Bard was an attempt by Google to enter the chatbot game.

• However, the initial reception of Google Bard was underwhelming compared to OpenAI's ChatGPT.
• Google's entry into the chatbot space sparked competition and innovation.

6. Microsoft's Bing chat leveraged OpenAI's GPT technology.

🥈86 09:57

Microsoft's Bing chat utilized OpenAI's GPT technology for enhanced chatbot capabilities.

• Bing chat quickly rolled out and gained traction due to its partnership with OpenAI.
• The integration of GPT technology improved the quality of chatbot interactions.

7. Control Nets brought new possibilities to AI art.

🥈83 11:37

Control Nets allowed for AI-generated images to be customized and posed.

• This advancement enabled users to create AI art with specific poses and positions.
• Control Nets expanded the creative potential of AI-generated images.

8. GPT-4 and Mid-Journey Version 5 showcased major advancements in language models and AI photos.

🥇91 14:02

GPT-4 and Mid-Journey Version 5 demonstrated significant improvements in language generation and AI photo realism.

• GPT-4 surpassed previous language models in terms of capabilities and performance.
• Mid-Journey Version 5 elevated the realism of AI-generated photos to a new level.

9. Improved video generation quality with Gen 2.

🥈85 14:32

Gen 2 videos showed significant improvement in quality compared to Model t875 scope, with no Shutterstock Watermark.

• Gen 2 videos were similar to Model t875 scope but had better quality.
• Examples of Gen 2 videos included realistic animations and cartoons.

10. July brought new abilities to chat GPT with code interpreter.

🥈88 14:58

Code interpreter in chat GPT allowed uploading and interpretation of PDF files, spreadsheets, and graphs.

• Code interpreter helped analyze data and create graphs.
• Anthropics clad 2 was also rolled out in July, providing a 100,000 token context window.

11. Llama 2 launch marked the rise of open source large language models.

🥇92 16:27

Llama 2, launched by Microsoft and Meta, was free for research and commercial use, leading to increased open source large language models.

• Llama 2's launch prompted people to switch from chat GPT to Claud for document summarization.
• Open source large language models gained popularity due to the availability of Llama 2.

12. Chat GPT gained new capabilities in September.

🥈89 18:50

Chat GPT gained the ability to see, hear, and speak, as well as vision capabilities for image analysis.

• Chat GPT could now have audio conversations and provide audio responses.
• Vision capabilities allowed image analysis and answering questions based on uploaded pictures.

13. Dolly 3 introduced highly prompt adherent AI image generation.

🥇91 19:47

Dolly 3, the most prompt adherent AI image generator, allowed complex prompts and conversational image generation.

• Dolly 3 could generate images with multiple elements based on complex prompts.
• It was integrated directly into chat GPT for easy image generation.

14. November saw the introduction of Grock and GPTs.

🥈87 21:14

Elon Musk introduced Grock, a large language model trained on X and Twitter data, and open AI announced GPTs for custom chatbots.

• Grock was known for its real-time updates and ability to engage in conversations.
• GPTs allowed developers to create custom chatbots tailored to specific niches or end goals.

15. December brought Gemini and Mixl of Experts.

🥈86 24:54

Gemini was announced, and Mixl of Experts introduced a new way of using large language models for chatbots.

• Gemini was made available in Google's Bard and offered a different experience than demonstrated in the video.
• Mixl of Experts used a router to send questions to specific language models tailored to the prompt.

16. 2023 was a year of rapid advancements in AI.

🥈85 27:41

2023 saw the fastest and most rapid advancements in AI, possibly in any technology ever.

• This year condensed what felt like 5 years of progress into one.
• The advancements in AI are expected to accelerate further in 2024.