The AI voice recorder market is booming: The hardware dreams of big manufacturers and the cost for users to abandon it

Recently, Feishu, in collaboration with Anker Innovations, launched an intelligent recording device on the market. ByteDance, the company behind Feishu, is responsible for the large model and software, while Anker Innovations is in charge of the hardware. The device is named "Recording Bean". It is indeed as small as a bean and can be clipped onto the collar, which is quite different from the mainstream magnetic recording cards on the market in terms of appearance.

Half a year ago, DingTalk's first intelligent hardware, the DingTalk A1 recording pen, was introduced, attracting widespread attention in the industry. Different from ByteDance, DingTalk decided to enter the hardware field on its own. In the following months, it seemed addicted to launching multiple hardware products for both AI to C and to B scenarios, which changed the outside world's perception of DingTalk.

Currently, there are already AI recording pen products from several companies in the market, such as the iFlytek series, the TicNote from Mobvoi, and the Plaud Note developed by the startup Plaud.ai. They look similar, have similar functions, and the price differences are not significant, seemingly indicating a homogeneous competition.

Actually, in the stage where the development of AI emphasizes implementation, it's not difficult to summarize the reasons why big companies like DingTalk and Feishu enter the AI recording niche market. We can easily list three reasons: finding a C - end entry for the implementation of large models; integrating the underlying ecosystem; and broadening the C - end monetization channels of large models through product sales and membership subscriptions...

The question here is, when looking at AI recording pens in the long - term development of the Internet, is this demand naturally generated by users, or is it that manufacturers are instilling "new demands" into users under the opportunity of AI? Further, in the context where there is a general lack of long - lasting products in the intelligent hardware field, what can determine the long - term trend of an AI recording pen?

Let's first recall an interesting phenomenon before and after the birth of the mobile Internet. Before smartphones became the main tool for accessing the Internet, users' commonly used hardware devices were diverse and each had its own function: cameras were used for taking photos, MP3 players or iPods for listening to music, professional recording pens for recording, laptops or netbooks for surfing the Internet, and Kindles for reading e - books...

Just as the opening sentence of Romance of the Three Kingdoms goes, "In the affairs of the world, long division leads to union, and long union leads to division." In the era of mobile Internet, smartphones have become the most important network entry point. Users gradually put aside those devices mentioned above, as they can find apps with similar functions in the mobile app store, which are also more portable. For example, taking photos once became a core competitive area for major mobile phone manufacturers.

Now that all mobile phones have a recording function, can AI recording be achieved solely through mobile phone software? Isn't the function of an external device just to pick up sound? Theoretically, it is possible, but it requires the protocol authorization of the mobile phone hardware. The intelligent recording pen bypasses this by connecting to the mobile phone via Bluetooth. Moreover, the recording pen has a longer recording distance, operates with low power consumption, and does not affect other operations on the mobile phone while recording. So, it is applicable in a wider range of scenarios.

More than a decade ago, due to work needs, I bought a Sony recording pen that used batteries. In terms of sound quality, it can still compete in today's market and has stable quality. At that time, some flagship - level recording pens were even used as players for high - resolution Hi - Fi audio sources. However, except for the sound quality, today's AI recording pens have comprehensively surpassed the old ones in almost all other aspects, as they are products of different eras. AI can convert sound into text in real - time, extend sound recording into knowledge storage, and the real - time translation function is like equipping an individual with a simultaneous interpretation specialist, enabling the underlying large model to perform in scenarios such as multi - language classrooms, business meetings, etc.

Therefore, the rapid implementation of large models at the consumer end is starting to challenge the status of the mobile phone as the special one. A series of hardware such as AI glasses, AI headphones, AI question - answering pens, and AI learning machines have emerged in the market, seemingly bringing us back to the era when multiple devices coexisted and each had its own function, except that many AI hardware still needs to be connected to the mobile phone.

Previously, users gradually lost interest in other devices as a natural selection process. Remember when Steve Jobs introduced the first - generation iPhone in 2007, the three icons of listening to music, making calls, and surfing the Internet continuously cycled on the large screen of the press conference and finally converged into one mobile phone. Now, users are expanding from using only mobile phones to multiple devices, which is more of a "newly created" demand by manufacturers within the AI framework. Users do have the demand to run and experience large models at the edge of the cloud - edge - end, and these demands are similar to a value - added service. This is also the logic behind why AI recording pens attract users to subscribe to paid memberships and unlock value - added functions.

For users, the cost of choosing and abandoning app software is very low. They can simply uninstall the app if they are not satisfied after using it. However, the price of AI hardware ranges from a few hundred to several thousand yuan, which raises the cost of abandonment for users, making them more cautious when making a choice. Currently, more than one intelligent recording pen manufacturer has stopped providing network services or membership support, resulting in the loss of users' audio files stored in the cloud space. For users who record important audio, this is a disastrous experience.

Therefore, recording files can serve as a knowledge base or evidence materials. The core requirement of users should be the long - term stability of the device and services, which reflects brand trust. Users hope that products can be finely polished, rather than pursuing novelty, coolness, and rapid iteration like other consumer - oriented intelligent hardware.

In this regard, the backing of big companies is a significant advantage for them to enter the AI recording field. However, there is also a problem. Platform - type companies, due to the characteristics of their business models, have always liked to pursue data volume and speed. This "war - report - style thinking" often conflicts with product thinking. Moreover, the long - term business status and horizontal authority of hardware within big companies may also affect the final product output.

Looking back at the history of big companies manufacturing hardware, around 2014 - 2016 was a key period. Products such as Amazon's Echo speaker, Alibaba's Tmall Genie, Apple's Apple Watch, and the AR glasses from Google and Microsoft all emerged during this time. Concepts such as smart home, metaverse, virtual reality/augmented reality were hyped up. Looking at it now, some concepts have quickly cooled down or been neglected, and the related hardware has also withdrawn from the stage and is no longer updated. Only a few products are still selling well.

Under the trend of Physical AI, big companies will manufacture hardware more frequently in the future, either through self - research or authorization, to accelerate the exploration of the boundaries of large models in the physical world. The Internet environment, which is used to grand narratives, likes to talk about the ecosystem right from the start. However, it's better to patiently wait for a good product that can "cross the cycle" and truly listen to the voices of users.

A person who claims to be deaf - mute sought recommendations for an intelligent recording pen on social media because they need to learn from mobile phone videos, but the videos have no subtitles. An AI recording pen can display text in real - time and store it. A good product shows care in the details.

This article is from the WeChat official account "One Thousand and Two Hundred Characters", written by keykey7 and published by 36Kr with permission.