According to Bloomberg, Netflix is beginning to test a search feature that utilizes OpenAI’s technology.
This new search tool will allow users to “search for shows with much more specific words, including their current mood, for instance, as stated by the company,” according to the report. This OpenAI-driven search will also enable users to ask questions that “extend well beyond genres or actor names.”
This feature, which is optional, is currently available for a select group of users in Australia and New Zealand on iOS devices.
Netflix representative MoMo Zhou confirmed to The Verge that Bloomberg’s report is correct. Zhou mentioned that the testing will expand to the US “in the upcoming weeks and months” and that there are no current plans for the feature to be available outside of iOS.
“It’s early days for the feature, and we’re really in a learning and feedback phase for this beta,” Zhou said.
OpenAI did not respond right away to a request for comment.
In a conversation on the Decoder podcast last year, Nilay Patel, editor-in-chief of The Verge, asked Netflix co-CEO Greg Peters how the company approaches AI. Here is part of his reply:
We have a long-established history of applying machine learning and artificial intelligence in our recommendation systems. We have been doing this for over two decades. Once again, we believe that it is our responsibility to proactively understand where technical innovations exist. How can we utilize that to benefit creators, enabling them to tell their stories more effectively, and also to provide our members with better user experiences?
Netflix is developing a new fundamental model.
This aligns with a research paper that Lamkhede and others recently published on the Netflix Tech Blog. In it, the authors detail the creation of a basic model for Netflix recommendations to manage the increasing complexity of numerous specialized models. Previous challenges included high upkeep costs, challenging innovation transfer between models, and limited time-based capture of user interactions.
Motivated by Large Language Models (LLMs), Netflix emphasizes a data-centric strategy featuring semi-supervised learning, as reported by the authors. The model analyzes extensive user interactions through a unique “tokenization” process, in which important events are recognized and supplemented with relevant context and content information. Unfortunately, the article does not clarify if this implies that an AI assesses users’ moods in the background or if users must explicitly communicate this to the AI through a prompt.
There are also early discussions regarding the new AI search on Reddit. The feedback so far has been quite critical, with many users highlighting previous limitations of the search, such as a lack of sufficient filtering options. One user expresses frustration in finding specific films from the 1980s using the previous search. Another user wishes for the ability to filter by release year or IMDb ratings. It remains uncertain whether such requirements can be addressed with the new basic model.
Netflix continues to innovate in the field of streaming advancements. The entertainment giant has begun exploring a new AI-enhanced search feature that employs OpenAI’s generative technology to help users discover content more intuitively and personalized. Initially made available to a few users in Australia and New Zealand, this functionality has the potential to transform how we engage with streaming services.
Here’s all you need to know about this groundbreaking development and its potential impact on future content exploration.
AI and Streaming: What’s the Latest on Netflix’s Search Feature?
Netflix’s latest initiative reveals a search tool powered by OpenAI, the team behind ChatGPT. Unlike traditional search systems that rely on specific keywords such as show titles, actor names, or categories, this refreshed feature allows users to search using natural language and emotion-driven prompts.
For example, instead of typing “romantic comedy,” a user could search for “something uplifting to watch after a challenging day,” and the AI will understand the emotional context and suggest relevant content. This marks a significant shift toward more human-like interactions with technology, enabling more fluid and emotionally resonant content discovery.
How Generative AI Enhances Content Exploration
Generative AI, such as OpenAI’s GPT models, is proficient in understanding and generating human-like text. When integrated into Netflix’s search capabilities, it allows the platform to interpret subtle user inputs and provide highly relevant results. This might entail:
Recommending shows based on mood-related descriptions like “uplifting,” “intense,” or “thought-provoking”
Understanding complex inquiries such as “a sci-fi series with a strong female lead in a dystopian environment”
Delivering results that align with a user’s desires rather than just matching keywords
This represents a significant improvement over conventional recommendation systems, which typically depend on past viewing habits and basic metadata.
Current Availability and Device Limitations
Currently, the AI-assisted search feature is available only to a limited group of users in Australia and New Zealand. Interestingly, Netflix has chosen to unveil this functionality solely on iOS devices, departing from its usual approach of testing new tools on Android first.
It is unclear if the next phase of testing will include support for Android devices or when the feature might expand to other regions, including the United States. However, given Netflix’s history of global expansions, it’s likely that broader trials will occur.
Netflix’s Continuing Involvement with Artificial Intelligence
This isn’t Netflix’s first foray into artificial intelligence. The streaming service already utilizes AI to:
- Analyze viewing habits and recommend personalized content
- Generate custom thumbnails based on anticipated user clicks
- Enhance streaming quality based on bandwidth and device capabilities
- Incorporating OpenAI’s generative abilities is a logical advancement of these efforts, aimed at improving user experiences to be more intuitive and emotionally engaging.
Future Ramifications for Streaming
The infusion of AI in content discovery may have far-reaching effects on the streaming industry:
- Improved User Engagement: By grasping user intentions and emotions, Netflix can provide more satisfying recommendations, keeping viewers engaged for extended periods.
- Competitive Edge: As streaming platforms compete for viewer attention, advanced AI functionalities could become a key differentiating factor.
- Ethical Issues: Like all AI applications, challenges related to data privacy, algorithmic bias, and transparency will require attention.
- This development aligns with wider technological trends, as companies increasingly utilize AI to create more personalized and adaptive user experiences.
Conclusion
Netflix’s AI-enhanced search function signifies a major step towards a more intelligent and emotionally attuned streaming experience. By leveraging OpenAI’s generative technology, Netflix is not merely refining content discovery; it is transforming the concept of searching in the digital age. Although it is still in its early stages, this feature could soon establish a new standard across platforms, setting a benchmark for user engagement within the entertainment sector.
As Netflix continues its innovative path, viewers can look forward to a future where finding the perfect show or movie is as easy as expressing their emotions.
Frequently Asked Questions
What is Netflix’s new AI-driven search function?
Netflix is testing a new search tool that employs OpenAI’s generative AI, allowing users to find content using natural language and emotion-driven queries, such as “something relaxing to watch” or “a suspenseful thriller with a twist.”
Where is this feature currently available?
This feature is currently being tested by a limited group of users in Australia and New Zealand, and it is presently only available on iOS devices.
Will the AI search function be available in the United States?
Netflix has not yet announced a timeline for a U.S. release, but reports suggest that the company plans to extend testing to more markets, including the United States, soon.
How does this differ from Netflix’s existing recommendation system?
Netflix’s traditional recommendation algorithm relies on viewing history and metadata. In contrast, the new AI-enhanced search interprets natural language and emotional signals, providing more personalized and contextually aware suggestions.
Is this feature accessible on Android devices?
Not at this time. Unlike previous Netflix feature rollouts that favored Android first, this function is currently exclusive to iOS, and it remains uncertain when or if support for Android will be introduced.
Does this imply Netflix is using ChatGPT?
While Netflix has not disclosed the specific model used, the feature is powered by OpenAI’s generative AI, which includes technologies similar to ChatGPT. This enables the search function to understand and respond to complex, conversational inquiries.
How does this affect user privacy?
As with any AI-driven features, data privacy remains a primary issue. Netflix will probably continue to follow its existing privacy protocols, but users should stay informed about how their data is utilized.