出品 | 51CTO技术栈（微信号：blog51cto）
Recently, it was rumored that Google was considering putting its latest AI search results behind a paywall.
As the absolute hegemon of the global search engine market, Google's every move is widely implicated. As we all know, the ecological landscape of the search field has not changed drastically for many years, but since last year, the popularity of large models and the rise of generative AI have stirred up a "pool of spring water".
In this context, the "leader" Google has also had to adjust its layout in terms of search. Now, Google's new move is still in the brewing stage, but its signals have already made some people start to question it.
1. Google: AI search is a hot potato
According to the Financial Times, Google is discussing internally whether to transfer some AI search functions to premium subscription services. When asked by a Google spokesperson for confirmation, the other party did not deny the news, but also did not disclose any details.
In fact, contrary to what most people expected, Google is in a dilemma when it comes to AI search.
While Microsoft's addition of generative AI capabilities to the Bing search engine hasn't significantly changed market share, the use of generative AI technology in search queries has the potential to change Google's existing advertising-focused revenue model.
Last year, Google launched the Search Generation Experience (SGE). Unlike Microsoft's Bing, which makes ChatGPT a "chat" interface, SGE combines search and generative AI, where users can get a separate summary "snapshot" of the answer and a link to a website that verifies the answer when searching. At the same time, Google will also clearly mark these responses as "Generative AI is experimental" and distinguish them from organic search results.
Of course, ads are also inserted next to AI-generated results. However, the shift from traditional search result lists to this new format has to be seen to wonder if this will change the way users interact with ads, which in turn will affect Google's revenue structure.
While SGE hasn't yet had a significant impact on Google's core search engine business, any potential impact on the company's advertising revenue aside, it could also be costly to roll out AI technology to more users due to its high dependence on resources. As a result, SGE is an advanced feature that is most likely to be set behind a paywall.
At the same time, a Google spokesperson also made it clear that they will actively develop more AI-related value-added services, but they have no plans to launch an ad-free search product.
Rosanne Kincaid-Smith, chief operating officer of the German company Northern Data Group, one of the world's largest providers of high-performance computing (HPC) solutions, expressed her displeasure after the news broke. In her opinion:
First, Google is in a unique position. Unlike tech companies like OpenAI, which focus on developing more new things, Google has a huge reach, reach, and a dominant position in the search engine market, and should make its AI search innovations public and "allow the masses to access and use the transformative technologies that AI brings."
Second, the essence of the discussion of placing AI search innovations behind paywalls actually involves AI ethics and data governance. "All of us in this field...... They have a direct responsibility to ensure that the ethical management of AI is an integral part of the service package they provide to the public, rather than something that requires additional payment. ”
2. The AI process of search: large manufacturers are irresistible and trapped by it
Regardless of the merits of this incident, a series of actions by observing Google can be seen that Google is essentially trying to find a way to make artificial intelligence and advertising business coexist. The main profit model cannot be changed, but it also has to continue to increase the AI search in terms of functionality.
"AI + Search" is never a new topic. But this trend has never been more irreversible than it is today.
With the continuous advancement of the AI-based process of search, AI search has become another "battleground" in the global business market, and new and old forces have shown their swords, whether they want to make cakes or share cakes, they are trying to unlock new scenes and find real killer features.
Abroad, Microsoft's combination punch made a strong attack, and Bing, which blessed GPT-4, quickly set off a storm; Google, which was stimulated, then launched the SGE service to integrate generative AI functions into search; and startups Perplexity AI and Glean were based on the C-end and B-end markets respectively, and grew into unicorns in a short period of time, and their strength should not be underestimated.
In China, there are also many heroes. Baidu relies on Wenxin Yiyan to upgrade its traditional search to an AI interactive search engine, Kunlun Wanwei's Tiangong AI search, 360 Group's 360 AI search and other products have also emerged; in addition, Taobao launched "Taobao Ask", and Douyin tested "AI search" in the APP, which provides a new solution for players in the fields of short video and e-commerce.
It seems that the competition in the AI search track has become fierce, but if you look at the larger margin, it is unlikely that AI search will disrupt the existing search ecology in the short term.
Image
Screenshot from StatCounter: From March 2023 to March 2024, the global search engine market share has basically not changed much, and Google is still riding the dust
First, there is limited innovation based on AI. So far, there hasn't been a real killer search product. Large manufacturers are relatively cautious when it comes to search innovation, and most mature search products don't have many amazing breakthroughs.
In addition, the accuracy of AI search needs to be improved. "Authenticity and reliability" is the core demand of users for search engines, but how to ensure that users see credible search results when AI-generated products are flooded on the Internet is still a problem that many enterprises need to solve.
Finally, the user inertia resistance is huge. Compared with AI search, traditional search has been rooted in most people's work and life for many years, and people are not motivated enough to switch from traditional search methods under path dependence.
Neeva, a niche search engine that had quickly gained a large number of users in a short period of time, ceased operations after only four years and was eventually acquired by Snowflake.
In a blog post announcing Neeva's closure, its founder wrote: "Throughout this journey, we've discovered that building a search engine is one thing, and convincing the average user to switch to a better option is another thing entirely. ”
3. Two routes: pay for the search itself, is it feasible?
Following Neeva's story, we can clearly see two completely different development paths around the search engine.
Route 1: Google Mode
Online advertising is still Google's main source of revenue. Among them, the position of the search advertising business is crucial.
According to Alphabet's financial report for the fourth quarter of fiscal year 2023, Google's advertising revenue in the fourth quarter was $65.517 billion, a year-on-year increase of 11%, of which the core search business revenue was $48.02 billion, a year-on-year increase of 13%.
Route 2: Anti-Google mode
Some emerging AI search companies are more inclined to return to the value of search itself, allowing users to see the content they like, rather than the content that has more "commercial value".
In terms of product form, the most important upgrade of AI search is that users finally don't have to look for answers in full-screen ads. This "anti-Google" way of searching determines that its main source of revenue will not be advertising.
Neeva, for example, was founded because its founder, Ramaswamy, believed that in the long run, an ad-based model would inevitably lead to deterioration in search results. To build a better search engine, you first need to change the incentives.
Similarly, emerging search engines Perplexity.ai and You.com also rely on subscriptions to monetize and make users pay for the search service itself.
Among the search entrepreneurs in China, Mita AI Search, known as the Chinese version of Perplexity, is also famous for its slogan of "no ads, direct results".
The screenshot is from the official website of Secret Tower AI
According to statistics from similarweb, as of March 1, the number of visits to the Secret Tower AI search website was as high as 4.783 million in the first 28 days. However, it is not yet clear what kind of business model Mita AI will adopt in the future.
Looking at the two routes on the whole, in terms of the development status, the advantages of the former are almost overwhelming. Perplexity, once regarded as a benchmark for entrepreneurship, is now also interested in turning the rudder.
In addition to the basic free version, Perplexity also offers a more powerful Pro subscription service. But this model is not enough to support its development. Perplexity has just announced that it is incorporating ads into its search results.
The screenshot is from the official website of Perplexity
Compared to the original ad-free model, this new model clearly opens up more revenue channels for the company. Analysts believe that Perplexity's decision to integrate advertising into its platform is not just a monetization consideration, but a strategic move to maintain and improve its services.
"Advertising gives them the resources they need to continuously improve their platform and stay ahead of the competition. However, this new feature raises more questions, such as how to seamlessly integrate ads into search results without compromising the quality and integrity that users expect. ”
4. Write at the end
It is an established fact that AI is involved in the field of search as a variable.
Let's take this as Google's new move as a way forward in its quest to commercialize AI search. Due to the natural contradiction between its interactive form and the advertising business, Google must step on a tightrope to make the two coexist as harmoniously as possible until a more suitable profit model is found.
And for emerging search startups, AI search is an opportunity to get search back to its roots: helping people find the page they want and avoid all obstacles along the way. However, how to truly survive the inherent pattern and achieve sustainable development is still the primary issue. At the same time, whether the dragon slayer boy will eventually become an evil dragon also needs to be vigilant.
People tend to overestimate the impact of new technology waves in the year ahead and underestimate the changes in the next five or even ten years. How can emerging search vendors fight their way through the traditional market? What will the real explosive AI search products look like? Will the global search ecosystem be reshuffled? There will be time to answer these unsolved mysteries.
Reference Links:
https://www.theregister.com/2024/04/08/google_search_paywall
https://www.thepaper.cn/newsDetail_forward_24783786
Hatps://vv.wusship.com/I/6009702.html
https://www.singlegrain.com/blog/n/perplexity-ads/
https://www.duydaima.com/group/topic/it/14562
To learn more about AIGC, please visit:
51CTO AI.x社区
Hatps://vv.51CTO.com/iG/
Source: 51CTO Technology Stack