OpenAI Unveils GPT-4o, Sparking Infinite Imaginations in the AI Search Engine Race
In the era of AI, the way we access information is undergoing a profound transformation.
In the early hours of May 14th, OpenAI announced the official launch of its latest AI powerhouse, GPT-4o, capable of achieving real-time video and voice interaction and set to be integrated into various OpenAI products in the coming weeks.
The release of GPT-4o once again shook the entire tech industry and sparked further contemplation on AI applications. With GPT-4o on the horizon, the direction OpenAI will take post-launch becomes a focal point of attention within the circle. Considering recent events in the AI domain, including Perplexity securing a new round of funding and ChatGPT transforming its homepage into a search bar, the prevailing consensus is that AI search is currently the most promising arena.
Search is one of the cornerstones of the internet infrastructure, serving as a core means of meeting rigid demands and nurturing “shelf-style” trading markets. However, with the advent of the AI era and its significant efficiency improvements, traditional search engines can no longer fully meet user needs. AI-powered search engines like Perplexity and New Bing have garnered widespread attention.
In 2023, traditional search engines generated $300 billion in revenue, capturing over 50% of internet traffic. Google, as a crucial gateway for internet information search, held a 90% share of global search volume. However, with the emergence of ChatGPT, Google’s search volume has gradually been overtaken by Bing. According to Statcounter data, in April 2024, Google’s search share stood at 80.74%, a decrease of around 6% from the same period in 2023.
A single stone has stirred a thousand waves. Faced with the escalating tide of AI, AI search engines are bound to become a battleground in the era of large-scale models.
Current Status Analysis of AI Search Engines
Since ChatGPT training data lacks real-time updates (GPT-4 data only updated to April 2023), its results exhibit a certain degree of lag. However, in the internet age where information updates occur in seconds, the inability to provide real-time data is a major flaw for ChatGPT. Hence, launching AI search engines is the way forward.
Moreover, with the explosion of information and the shifting search logic of users from pursuing comprehensive information to “discerning the true from the false” amidst the sea of information, AI’s ability shines in this domain.
AI search engines address the issue of result lag by incorporating real-time data, fine-tuning with large models, and integrating the latest network data and information. They can simultaneously balance depth and efficiency, providing users with more precise information in natural language format. For contemporary internet users submerged in the ocean of information, this is undoubtedly a much-needed lifeline.
Currently, there are several general AI search engine products, mainly divided into Web2 and Web3 tracks. Products like Perplexity, Microsoft Copilot, YouChat, Brave Search, Andi, Komo, ThinkANY, Phind, iASK.ai, Adot, PreSearch, TypoX, QnA3, and Kaito fall into this category.
Perplexity AI
Perplexity is arguably the most influential Web2 AI search engine, built on OpenAI and API’s GPT, supporting online functionality and free use of GPT-4 features. It provides search results through dialogue with users, with the ability to switch models and unlock features like image generation, significantly improving model performance and accuracy. However, for Web3 users, its data sources are not comprehensive enough and cannot integrate on-chain data.
At the beginning of 2024, Perplexity AI secured a record-breaking $73.6 million investment, with a valuation reaching $520 million, backed by investors such as Nvidia and Jeff Bezos. In early March, Perplexity AI raised around $63 million at a valuation of $1 billion. On April 24th, according to sources, Perplexity is undergoing a new round of financing of at least $250 million, with a valuation potentially reaching $2.5 to $3 billion.
Microsoft Copilot
Microsoft Copilot is Microsoft Bing’s AI search engine, accessing vast amounts of data and generating answers based on facts and objective data. It has very fast response times and supports multiple languages. However, its reference data is not comprehensive enough, and it lacks creativity, analytical, and summarization capabilities.
YouChat
YouChat can search real-time data and provide quick responses, delivering accurate information and insights. However, it cannot understand certain complex contexts or emotional backgrounds, leading to the inability to provide entirely appropriate answers.
Brave Search
Developed by Brave Software, Brave Search is a relatively new AI search engine aimed at providing a more privacy-protected and secure search experience. One of Brave Search’s features is its focus on user privacy protection and ad substitution models. However, being a newer AI search engine, Brave Search may still require development and improvement in the breadth and depth of search results.
TypoX AI
TypoX AI provides an intuitive Web3 knowledge search and learning experience, integrating fragmented information sources and Web3 protocols on one platform, effectively breaking down the barriers to entry into Web3. It features user-friendly interaction, rich on-chain and off-chain data, and supports multiple languages. However, its functionality is relatively single, unable to address the illusion problem of AI models.
Clab.ai
Clab.ai’s main feature lies in being a Web3 vertical AI product, capable of handling large amounts of non-standard data, thereby expanding the sources of data. Additionally, it includes a significant amount of information from Twitter. According to official information, Clab’s data is updated on a minute-by-minute basis, a significant feature, and ensures improved result accuracy through cross-verified information.
Clab.ai was launched in 2023, with the official X ceasing updates in November 2023, and failing to launch a public beta as promised.
Kaito
Kaito enables users to search Twitter, Discord, Mirror, Medium, and other mainstream Web3 information channels with one-click. Additionally, Kaito ranks results and provides personalized recommendations, supporting image and voice transcription to integrate information into content. However, the reliability of data transcribed through secondary processing may be relatively lower, requiring cross-verification from other data sources.
AI Search Engines: A Necessity for Web3
We are in an era of information explosion, where Web3 alone generates vast amounts of information daily. Many people spend hours each day browsing various platforms in search of the latest news, hoping to find early investment opportunities or valuable information. However, amidst this flood of information, 99% of it comprises false, hype-driven, or junk information, wasting time and energy and significantly reducing our efficiency and returns.
For ordinary investors and users, obtaining crucial information from this requires a lot of time and effort. Although we are inundated in the ocean of information, information hunger still plagues us. Creating a comprehensive Web3 AI search engine is an urgent key issue to address.
While Web2 AI search engines are the battleground for giants like Google, OpenAI, and Perplexity, Web3 vertical AI search engines still lack widely adopted cases. Projects have begun to reconstruct Web3 vertical AI search engines, such as Adot, TypoX AI, PreSearch, Kaito, but there is still a long way to go before widespread adoption.
AI has made remarkable progress in the Web3 domain, with CoinGecko data showing that the total market value of 139 AI sector projects has reached $27.5 billion, often leading the crypto market. Conservatively estimated, by 2030, the Web3 AI market has the potential to grow to tens of billions of dollars. There is undoubtedly a place for Web3 AI search engines within this growth trajectory.
In the wave of the AI era, our opportunities as ordinary people lie in seizing and following early projects, especially the emerging Web3 AI search engine track. In the wild west era, the chances of encountering Alpha are even higher!
Free AI Search Engine — Adot
Adot, a Web3 AI search engine, is currently available for free use. It effectively addresses the illusion problem of AI models, integrating multiple valid data sources to provide efficient and accurate responses to users. Adot has introduced the innovative concept of “Search to Earn,” enabling users to earn rewards through searching, making it a highly noteworthy project in the Web3 track!
On April 25, 2024, Adot announced the successful completion of its Pre-A round of financing, totaling $3 million. This round of investment was led by Hash Global, with gCC (gumi Cryptos Capital) and Liang Xinjun’s family office participating.
To date, Adot has raised a total of $6 million, with other investors including SevenX Ventures, Mask Network, Var Capital, MixMarvel, among others. These funds will be used to advance the development of the decentralized AI search engine.
Adot was founded by former Google engineers and AI experts. Adot’s founder, Dr. Wei, is a former Google core engineer and AI expert with over 20 years of experience in search engines and AI technology.
Adot’s core team also boasts extensive industry experience, with each member averaging over 10 years of experience in the search engine and AI fields. Its product partners have held key product positions at Meituan, Ctrip, and an AI company acquired by ByteDance; the search algorithm leader holds a master’s degree from Peking University and has 15 years of search engine experience; the backend technology leader has built platform systems with over 200 million users from scratch.
Adot’s vision is to create a decentralized search network that surpasses traditional search engines, opening up all high-quality data to every user and developer. Users can locate accurate, scalable, and indexed data within seconds and earn rewards for their contributions to data sources. Developers can build their own search engines and recommendation systems based on Adot’s SDK and API.
Currently, users can earn Agems points rewards on the Adot (a.xyz) platform through daily check-ins, daily searches, and inviting friends. Agems can be used to participate in lottery activities, redeem prizes, and exchange platform tokens at a certain ratio, among other uses.
Conclusion
For Web3 users, using AI search engines can save 90% of their time. This saved time can be used for learning, entertainment, fitness, etc., improving the quality of life while enhancing investment returns.
A new track represents new opportunities. The decentralization and community-driven nature of Web3 allow early participants to often receive decent returns through personal contributions and involvement. The Web3 + AI search engine track remains an undeveloped virgin territory, worthy of close attention for its development and evolution, to seize early opportunities and dividends!
Author: 小树苗(https://twitter.com/chloe95117)
Translator: Adot Team
About Adot
Adot is a decentralized search network for AI Internet. It not only provides users with a more convenient and intelligent Web3 content search experience but also assists developers in quickly building their own personalized search functionalities.
Adot’s mission is to build a decentralized search network that all-round surpasses traditional search engines to make all high-quality data openly accessible to every user and developer.
Website: https://www.a.xyz (yes, we’ve got the coolest Web3 domain out there!)
Product: https://ai.a.xyz
Twitter: https://twitter.com/Adot_web3
Telegram: https://t.me/AdotWeb3_Official
Medium: https://medium.com/@adotweb3
Github: https://github.com/adotweb3
Waitlist: https://forms.gle/9voMawXDQXyUp2uy6