According to the first global GenAI landscape, around half of the countries that invest in AI create their own generative models

The first volume of the Global Generative AI Landscape 2024 has been created by AIport, an online community devoted to reporting on the most recent advances in machine learning worldwide. This first edition looks at prominent GenAI participants from across the globe in a number of important areas.

This is the first published examination of a generative AI landscape that has an emphasis on regional characteristics and covers four times as many countries as the typical GenAI landscape.
According to Tortoise’s Global AI Index, every one of the 62 nations with investments in the AI sector was examined throughout the study process.

The group of editors and data scientists first selected the in-house model creators, who were then sorted into ten GenAI categories after cross-referencing them with the most recent GenAI landscapes from Sequoia Capital, Antler, Base10, and other sources. Lastly, the data was categorized into the following continents: Africa, Europe, Asia, Oceania, South America, and North America.
The goal of AIport’s first book on the global GenAI environment is to provide a fair picture of multinational corporations, including both Western and non-Western businesses. The landscape provides a thorough examination, outlining the participants involved in the development of GenAI solutions, their locations, and the particulars of their contributions. A total of 128 generative models from 107 businesses are included in it.
“We noticed that many generative AI landscapes tend to focus either on the Silicon Valley giants or the tech powerhouses of Europe, covering no more than 10 countries on average,” said Avi Chawla, a data scientist and community manager at AIport. Although this method accomplishes its goal, it is not truly able to provide a whole picture. We made the decision to look into this further, and after weeks of investigation, this is what we discovered. We think that Volume 1 of our Global Generative AI Landscape 2024 offers a genuinely global perspective. Additionally, we want to investigate additional facets of GenAI in greater detail in the future.
The scenery and main features

Just 35 of the 62 nations included in the Global AI Index create their own GenAI solutions internally. About 90% of them concentrate on a single model type.
North America (USA), South America (Argentina), Europe (UK and France), Asia (China and Israel), Oceania (Australia and New Zealand), and Africa (South Africa) are the regions with the highest number of operational GenAI enterprises.
North America has the largest average number of GenAI models per firm and is the only area with at least one model from each of the ten model categories.
Multimodality has been included in GenAI models by around 10% of all the businesses in the research; most of the developers of these models are based in the United States. This suggests that while multimodality is a new trend, its uptake outside of North America is still in its infancy.
Eleven businesses worldwide have created many GenAI model types. With five different GenAI model types (image, video, audio, 3D, and code), Stability AI leads the field. OpenAI and Google, with four model types each, are closely behind with chatbot, audio, video, and multimodal, and text, image, audio, and multimodal, respectively.
Companies like Microsoft, Meta, Tencent, Baidu, and Yandex are a few that have created two or three different kinds of GenAI models.
Thirteen businesses have created many models in a single GenAI category. MosaicML has two versions of its MPT for code creation, AssemblyAI offers two speech-to-text models, and IPOXCap has two chatbots for business intelligence applications.
The goal of AIport, an online community of data scientists, authors, and researchers, is to provide a global viewpoint on artificial intelligence. Realizing that the majority of publications on machine learning mostly concentrate on the “big leagues” in the West, AIport aims to be more inclusive by extending the storyline and changing the perspective. This strategy guarantees a more unbiased and diversified representation, providing a comprehensive perspective on the worldwide AI community.