To the extent generative artificial intelligence poses a disruptive threat to existing industries and firms, it will do so to the extent that GenAI is a content-generation tool, and will therefore pose the gravest strategic threat to content-producing industries.
Film and TV, music, question-and-answer sites and search are among the obvious flashpoints. To the extent that people use search to find answers to questions, GenAI could disrupt the business model by reducing advertising inventory, for example.
There is, for example, some threat to Reddit user volume if people can use GenAI instead, the same reason GenAI is a potential threat to Google search, a replacement for some amount of code writing, acting, scriptwriting, video effects and music creation.
“In this work, we document a reduction in activity on Stack Overflow coinciding with the release of ChatGPT, a popular LLM,” say researchers. “Within six months of ChatGPT’s release, activity on Stack Overflow decreased by 25 percent relative to its Russian and Chinese counterparts, where access to ChatGPT is limited, and to similar forums for mathematics, where ChatGPT is less capable.”
In other words, people reduced their content contributions on the question-and-answer site. The researchers believe alternative methods--using ChatGPT--were used as a substitute.
Ironically, GenAI might also affect the amount of user-generated content that is created. That could, in principle, also affect social media business models to an extent. If people can find answers or create content in ways that do not require user generation, there could well be less user-generated content.
How that affects content businesses built on user-generated content is unclear, both in scope and scale. Many observers focus mostly on the threat of job substitution. That is almost certain to happen, to some extent, but might not actually be harmful for industry business models, even if it reduces the need for labor in some industries.
Reduced ad revenue is the biggest threat for many content-based apps that are reliant on advertising revenue for monetization. That seems obvious for search providers, for example.
Decreased page views and clicks mean reduced ad impressions and clicks. Also, AI-generated answers should mean many users are less likely to click on ads or organic results.
On the other hand, an increase in ad inventory due to AI-generated content could lead to lower cost-per-click rates over time, and possibly affect profit margins.
To the extent that ad-supported apps rely on volume, organic traffic could drop if GenAI becomes a major substitute. Websites could see a significant drop in organic traffic if GenAI becomes a viable substitute for users seeking answers and content.
As users rely more on AI-generated answers, they may visit fewer individual websites directly or through referrals.
With quick AI-generated answers, users may spend less time on individual sites, potentially impacting metrics like time on site and pages per session. So less user engagement is another potential business issue.
Providers of search engine optimization might find their algorithms are less valuable when direct answers to queries are possible, without referrals to websites that might provide answers.
Commoditization of basic information also seems likely, as websites offering basic information lose their value.
The big shift in potential value is the change in behavior from “search and browse” to “ask a question.” The former drove traffic to content generator sites; the latter eliminates those visits. So even as search provider revenue models are potentially disrupted, so are internet monetization opportunities for virtually all content providers.
No comments:
Post a Comment