I don't know when it started, the hot spot in the field of rubbing technology has become the default operation of the entire advertising and marketing industry. At that time, the concept of VR just appeared in the mass market, and various industries have planned a series of "pies" that have not yet been realized at a very fast speed, including VR education, VR training, VR art creation, etc., and then concepts such as blockchain and NFT have become cash cows in the field of Internet finance, and the meta-universe concept that is still in its infancy has also successfully joined hands with the "real estate" industry to cut a wave of "Internet leeks".
Image source: OpenAI
And this year's AI big model, which exploded with ChatGPT, will naturally not be let go by the marketing industry. Let's not mention the AI models of the Internet giants, ChatGPT alone has been integrated into products in all walks of life in less than half a year:
Some car brands use AI models to improve the dialogue quality of smart cars, some game companies use AI models to enrich dialogue texts in online games, and even some companies add AI models to smart home appliances to meet the needs of some users chatting with refrigerators, air conditioners, and kettles when they are idle. Although Xiaolei does not understand whether this "Q treasure" behavior can be converted into actual sales, considering that smart home appliances are often the interactive entrance of smart home and whole house intelligence, such a quick application seems to make sense.
But after seeing the news that a brand plans to launch a "ChatGPT bike", Xiaolei was still shocked by the "creativity" of these brands.
Can ChatGPT ride a bike?
Although the brand called the bike a "thinking bike" on the product introduction page, after a simple search, Xiaolei found that this ChatGPT bicycle was similar to another electric bicycle of the brand in terms of appearance, configuration and technical indicators, and could even be said to be completely consistent. Of course, since the "original" bicycle itself is also a so-called smart electric bicycle, which can be connected to a mobile phone and synchronize positioning and data, Xiaolei is not good to say whether this "ChatGPT bicycle" is a new model upgraded or an updated new version.
In addition, the brand has not announced whether this ChatGPT model is built into the electric bicycle machine local, or built into the mobile phone App, on the brand's official website we can't even find this so advanced electric bicycle, as for the built-in ChatGPT version we naturally have no way of knowing.
Image credit: Urtopia
According to the brand, this "ChatGPT bike" can "easily request navigation instructions and have a natural conversation while driving", with ChatGPT, users can enjoy an immersive and interactive riding experience, just like talking to the assistant on the bike.
But the problem is that the electric bicycle brand is mainly for European and American users, and the "assistant on the car" is generally Google Assistant and Siri. If I really want to talk to them, why don't I directly use the Bluetooth headset to wake up the voice assistant in my phone?
Of course, the brand can also use AI technology to analyze data such as users' riding habits, so that the vehicle's power curve matches the user's riding preferences and improves the user's riding experience. But to be honest, these features have appeared in domestic brand electric bicycles before ChatGPT came out, and this deliberate bundling of ChatGPT seems like a one-sided rubbing hot spot anyway.
Products need heat, and OpenAI needs heat
In order to understand the opinions of the "parties", Xiaolei also chatted with ChatGPT to understand the views of AI large models on this hot behavior. As a qualified "Duanshui master", ChatGPT naturally did not give a clear answer, saying that "it does not mean that everyone will agree with or support this strategy"; But at the same time, ChatGPT also believes that "this is a more common trend".
After all, from the original goal, one of the purposes of creating AI is to liberate human productivity and at the same time use AI's powerful computing power to deal with problems involving complex data. Taking AI language models such as ChatGPT as an example, the blessing of AI can solve the pain point that traditional voice assistants cannot understand fuzzy instructions, making voice control and interaction more natural and user experience better.
Image source: OpenAI
On the other hand, introducing language models such as ChatGPT into different categories of products is actually what AI companies such as OpenAI want. In the case of ChatGPT, for example, OpenAI has prepared full third-party development support for it, allowing and expecting third-party developers to integrate ChatGPT into their services. After all, training an AI model requires a lot of user data, and the introduction of third-party forces can in turn support the development of AI models, which is extremely important for OpenAI and any AI model.
Besides, in addition to the popularity of third-party products that need to rely on ChatGPT, ChatGPT also needs these third-party services to maintain its popularity.
According to the monitoring of data website SimilarWeb, the number of ChatGPT visits has fallen sharply since entering the second quarter, and there may even be negative month-on-month growth in June. According to SimilarWeb's data, in the first five months of this year, ChatGPT visits increased by 131.6%, 62.5%, 55.8%, 12.6% and 2.8% respectively, and has been in a downward trend.
Image source: AppStore
In order to maintain the number of visits, OpenAI not only relaxed the original IP restrictions, but also launched a mobile app to lower the threshold for user access. But as more and more Internet giants join the battlefield of AI models, how much influence is left of OpenAI?
It is true that Xiaolei is well aware that conversational AI has extremely strong strength, and GPT-4 is an AI model with almost no opponents at this stage. It is a "tool" for many people, and if other companies can make good use of it, there is no doubt that conversational AI will be a good helper to improve work efficiency and greatly enrich the user's interactive experience.
However, there are currently very few categories and brands that can really use AI, and most of the products that "join" ChatGPT are just "adding gimmicks" with the mentality of Tuyile. Xiaolei also understands that the popularization of AI is inseparable from these "blind" attempts, but when he sees these products that pull ChatGPT when there is nothing to do, Xiaolei always remembers the way the whole industry tried to rub blockchain hotspots at that time. For these concept products, a simple "next time" is the best evaluation of their short life cycle.