ChatGPT and other language model AIs from companies based out of the United States have seemingly dominated the niche thus far. In response, the Chinese companies Alibaba and Baidu launched their respective versions of AI chatbots—Tongyi Qianwen and Ernie—to enter the emerging market in artificial intelligence. With more countries and companies noticing the potential of language model artificial intelligence, Russia, too, has reportedly been working on its own edition of a chatbot. Sberbank—the country’s state-owned financial entity—recently announced its foray into the AI space with GigaChat. The chatbot is being touted as Russia’s formal contribution to the global AI boom. As the country faces increasingly hostile relations with its Western counterparts following the outbreak of conflict in Ukraine, the country has increasingly looked inward for production and self-reliance. Sberbank, in its initiative, identified AI to be among the key areas for self-reliance, given that Western inflows—both financial and technological—have virtually halted for the Russian Federation. Notably, Gigachat’s launch also comes at a time when an another Russian firm—Sistemma—dedicated to data analysis and processing, recently launched another competitor to ChatGPT titled “SistemmaGPT” with claims to encyclopedic knowledge and seamless communication capabilities in Russian.
The market of AI and interactive chatbots grows more competitive by the day, and an increasing number of countries intend on getting out their iterations to the global market before monopolies are established. OpenAI’s launch of GPT-4 further created a sense of urgency among competitors, with several companies naming and launching their editions of artificial intelligence in its immediate aftermath. An analysis of existing trends suggests that industry rivalries are bound to transcend the corporate sector, leading to global competition. Though GigaChat’s technical details are yet to be unveiled entirely, the launch holds numerous implications for the rapidly growing industry and people impacted by it.
Russian Artificial Intelligence: A Background of GigaChat’s Launch
GigaChat has so far been marketed as an alternative to ChatGPT in Russia. However, the launch of the Russian chatbot comes at a rather interesting time, given that the country faces several challenges including war and diplomatic isolation. Sberbank has reportedly been investing heavily in technology to further the country’s capabilities in the tech space. Russian industries are also not entirely unaware of the workings of AI. Moreover, the nation has an existing reputation for creating complex software. Russia has also been accused of protracted hacking campaigns against its rivals while deploying automated technologies. With these apparent realities in the fore, it comes as no surprise that Sberbank looks to branch out in the generative AI niche to bolster the nation’s capabilities in what is a current global phenomenon.
Russia had banned ChatGPT soon after its launch, fearing subterfuge through the application by its opponents. Subsequently, OpenAI, too, blocked the usage of ChatGPT in Russia, leaving the market in the nation untapped and ripe for the picking. The country is looking to closely monitor and regulate the generative AI market due to growing concerns surrounding dissent and external manipulation. Hurried releases from other American firms such as Google and Hugging Face further created pressure on the international market to come up with a response. As growing concerns surrounding a renewed technological race between Russia and the United States took root, Sberbank launched GigaChat in the last week of April 2023. In what is an ongoing AI gold rush, Russia’s entry into the language model space further hints at global multipolar competition in technology’s most popular subsector.
An Overview of GigaChat’s Capabilities
GigaChat is being developed by Sberbank’s technical divisions—SberDevices and SberAI. The chatbot is based on the Neural Omnimodal Network with Knowledge Awareness (NeONKA) model. Its architecture involves various neural networks working in unison, bolstered by a human feedback mechanism to improve learning outcomes for the AI. The interface of GigaChat is purportedly multimodal, providing it an edge over GPT-3.5, which is the current model that ChatGPT uses. Sber has also integrated another language processing model to provide the chatbot with an image generation capability, making it a possible competitor to other extant image generation platforms like Midjourney and Dall-E. The initial release of GigaChat and its NeONKA model might involve over 13 billion parameters and will first be evaluated by a small community of testers that will gain access to the chatbot through a closed Telegram group. Further development will then proceed based on the testers’ feedback before the model is released to a broader open-source community that can add consistent improvements and patches to the AI’s source code.
GigaChat is said to be more fluent and comfortable in Russian as opposed to English and is meant to serve the country’s domestic population. Apart from image generation and processing, GigaChat can answer questions, provide fact-based responses, generate code for programs, and even develop rudimentary software. Other AI developers like Google and OpenAI have also tapped into the AI coding market with their respective offerings such as Google Codey and ChatGPT’s Code Interpreter. Gigachat might have to compete with tools that have a better framework and access to resources on the programming front. Moreover, despite its impressive feats, the chatbot is supposedly not as efficient at holding long-form conversations with its users, unlike its other counterparts. However, this feature might be scaled up over time as the bot is released to an open-source community for consistent improvement. GigaChat’s launch might inspire other local players to step in and contribute to the country’s foray into the AI, machine learning, and big data domains while further enhancing Russian presence in the industry.
OpenAI Competitors and the Future of Chatbots
Clearly, OpenAI’s GPT series of chatbots have caused nothing short of panic and urgent responses from both American companies as well as international players. As firms and countries continue to develop advanced forms of artificial intelligence to support scientific, business, and generic requirements, close monitoring and a commitment to responsible AI creation must be ratified. Despite the extant regulations, the impact of chatbots on sensitive societal components like academic institutions must be studied. As witnessed in the immediate aftermath of ChatGPT’s launch, threats to academic integrity and student learning outcomes cannot be compromised for rapid technological progress. Despite the rivalries and pitched competition, companies and countries must arrive at a consensus to support mutually beneficial policies and the creation of artificial intelligence that adds value to all aspects of modern human subsistence.
FAQs
1. Which company launched Gigachat?
Sberbank—a Russian financial services firm—launched Gigachat in April 2023 as a competitor to growing global AI chatbots like ChatGPT. Gigachat is capable of performing various tasks like image generation, AI coding, AI writing, and compiling information.
2. What language model does Gigachat use?
Gigachat is based on a versatile language model that combines together various neural networks to function. Known as the Neural Omnimodal Network with Knowledge Awareness (NeONKA) model, Gigachat draws its information and data from this model and carries out various tasks based on user prompts.
3. Is ChatGPT used in Russia?
Russia banned ChatGPT soon after its release fearing security risks. OpenAI, too, banned the platform in Russia citing similar concerns. Ever since, Russia has developed its own chatbots like Gigachat and SistemmaGPT for its local populations.