Artificial intelligence has made significant progress in the past month or so. The release of ChatGPT last year has spurred both its developers and competitors to come up with new solutions. The changes affect language models, visual models, and research in general.

One of the notable developments is the release of the latest iteration of OpenAI's language model, GPT-4. GPT-4 promises to be even more powerful than its predecessor, capable of understanding and generating more complex language. The new model better understands long-term dependencies and conceptual relationships and, unlike its predecessors, is multimodal, meaning it can handle both verbal and visual inputs and outputs. A research team at Microsoft even claims that GPT-4 may exhibit the first signs of general intelligence and can be compared to human understanding.[1]

Microsoft, as a significant investor in OpenAI, has embraced the opportunity and integrated the new technology into its software solutions. The Bing search engine now provides answers generated by artificial intelligence in addition to search results, and the AI helps better understand user queries and provide more relevant search results. Copilot, a programming assistant based on artificial intelligence developed in collaboration with OpenAI, has also been enhanced and now offers developers more accurate and efficient coding suggestions.

The recent advancements by OpenAI and Microsoft have spurred other technology companies to compete and come up with their own solutions. Google introduced its chatbot Bard in February, which received a lukewarm response.[2] Similarly, the experiment by the Chinese technology giant Baidu faced criticism.[3] Both companies’ stocks dropped significantly after the unveiling of their new products. While the market for virtual assistants based on large language models is still dominated by OpenAI, competition is expected to intensify in the near future. With ChatGPT becoming the fastest-growing software product in history (reaching 100 million users in just two months), competitors cannot afford to lag behind.

However, AI news is not limited to language models alone. Two giants in the visual industry, Nvidia (a graphics card manufacturer) and Adobe (the maker of Photoshop), have joined forces to create and integrate new solutions based on artificial intelligence. On one hand, the aim of the new models is to simplify the work of human creators, while on the other hand, they offer competition to generative visual models such as Stable Diffusion, Dall-E, and Midjourney.[4]

A significant milestone in the world of visual artificial intelligence is the second-generation video generator developed by the American company Runway ML. While the previous solution could animate images, the new technology allows for the generation of short videos purely based on text.[5]

However, the AI revolution is not solely driven by private companies. In February, Stanford University unveiled its Alpaca model, a humorous reference to OpenAI's language model LLaMa, which serves as the foundation for Alpaca. Alpaca is a chatbot with similar functionalities to ChatGPT and, in some cases, is almost on par with OpenAI's product but requires significantly less resources and funding. Although Stanford had to remove the public demo version of the model from the internet due to financial and security considerations, Alpaca is a significant step toward the accessibility and democratization of artificial intelligence.

Lastly, Stanford University's Artificial Intelligence Lab has adopted a new artificial intelligence model called Alpaca, which can generate realistic and personalized responses. What sets Alpaca apart from other AI models is its ability to produce multiple responses in a short amount of time, allowing for more natural and human-like conversations. This technology can be used to improve conversational robots, virtual assistants, and even video game NPCs. These advancements are paving the way for an AI revolution in various industries, from healthcare services to entertainment. The societal opportunities and challenges that come with it are still uncertain. What is clear, however, is that artificial intelligence will increasingly play a bigger role in our everyday lives.

1. [2303.12712] Sparks of Artificial General Intelligence: Early experiments with GPT-4 (arxiv.org); Microsoft Now Claims GPT-4 Shows 'Sparks' of General Intelligence (vice.com)

2. Alphabet Stock Drops 8% After Google Rollout of AI Search Features (wsj.com)

3. Baidu moves Ernie chatbot launch behind closed doors (techmonitor.ai)

4. Nvidia launches new AI service - and Adobe is on board | TechRadar

5. Gen-2 by Runway (runwayml.com)