Will AI replace Blinkist? The answer is a resounding no, and our SVP of Product + Content, Victoria Busse is here to tell you how we’re so sure.
Victoria Busse — August 23, 2023
As the learning and development landscape continues to evolve, the role of technology cannot be overstated. We’re just entering a new era of advanced technologies fueled by advances in generative AI. Already, we witness its transformative impact in fields like creative, advertising, healthcare diagnostics, and manufacturing efficiency. And we can expect this innovation to leave a lasting mark on learning and development.
In a playfully ironic twist, I’d like to address a common concern that we’ve been hearing frequently since the excitement of Chat-GPT started:
The introduction of artificial intelligence is not the end of Blinkist, but rather the beginning of a remarkable new chapter for us, one that will help us amplify rather than replace human learning.
As Blinkist’s Product & Content Leader, I am looking forward to sharing our journey of integrating AI into Blinkist, putting any concerns to rest that AI could be the demise of Blinkist, and discussing its incredible potential.
Before delving into our Generative AI endeavors, it’s important to understand where we began. Over a decade ago, Blinkist started as an educational app aimed at providing people with concise yet insightful takeaways from the best non-fiction books, podcasts, and experts across various fields.
Rooted in our mission to foster continuous learning, we aimed to make learning faster, more convenient, and remarkably efficient.
Recognizing the recent emergence and enthusiastic reception of ChatGPT, we realized the importance of adapting to current trends and embracing Generative AI to elevate our product and expand our content offerings.
It’s worth noting, though, that we have been utilizing Machine Learning for recommendation services, a step that naturally led us to explore the revolutionary potential of AI, particularly in the realm of Generative AI.
Creating Blinks is a manual and time-consuming process. It involves several steps and people, such as:
As you can imagine, this process requires a lot of human resources, skills, and expertise. It also limits the amount of content we can produce and offer to our learners.
With AI, we wanted to explore and find a way to scale our content production while preserving our quality and retaining our human touch.
We started exploring Generative AI as a possible solution for our content creation challenges in 2022. We were inspired by some of the advances and breakthroughs in natural language processing (NLP), which is the ability of machines to understand and generate human language, and the emergence of GPT-3, a large language model created by OpenAI that can produce text on almost any topic based on a given prompt.
These developments sparked our curiosity. Could we use GPT-3 or similar models to help us create Blinks faster and easier? Could we use advances in voice generation to produce Blinks at scale? Could we use AI to further personalize our content for our learners based on their preferences, goals, and needs?
Armed with these questions, we embarked on an experimental journey to harness the potential of AI tools.
At Blinkist, we currently use and explore Generative AI for various purposes, ranging from content creation and production to improving content personalization. Here are two ways we are currently using AI:
We use AI tools like GPT-4 to help us:
Additionally, we are utilizing tools like Elevenlabs for voice generation and audio content production, capitalizing on rapid advancements in voice cloning and generation to enhance our audio offerings.
Take a listen below for an inside look at human vs AI. Can you spot the differences?
However, we are not relying on AI alone to create our Blinks. While AI’s capabilities are truly remarkable, there are still certain areas in which it cannot be entirely relied upon. For example, AI can hallucinate at times, providing information which isn’t accurate. Therefore, human producers and editors continue to play a pivotal role in briefing, reviewing, editing, and approving the AI-generated Blinks. They further add their own voice and style to make our Blinks more appealing and relatable for our learners.
While AI-generated narration and audio mastering capabilities continually advance, they cannot yet compete with human quality and consistency. AI-narration, for example, has difficulty matching the intonations, cadence, and emphasis of human speech. When we want to convey an important message verbally, we slow down our speech and emphasize the keywords and phrases. This helps us make the message clear and stick in the listener’s head.
AI-narration and audio editing cannot match this to the same degree —at least right now. Therefore, it takes more mental bandwidth and focus for a listener to understand and remember the core message in automated speech. This makes passive listening nearly impossible. So, when our team of audio experts work with AI, they ensure that our audio content remains high-quality and engaging to listen to.
Human creativity and judgment are and continue to be essential for producing high-quality learning content. This is something that AI won’t change.
AI, however, will become a catalyst, amplifying what we can achieve. It will help us unlock new levels of efficiency and scalability, all while keeping the unique touch of human expertise that defines and has defined Blinkist for over a decade. Think of it like a perfect blend of human understanding and AI’s smarts, leading us into a new era of incredible learning adventures.
It may be all the rage right now, but we have been using machine learning and Generative AI for years to help users discover the most relevant content.
For example, we use Machine Learning (ML) models to recommend content that aligns with individual interests and behaviors, Generative AI to determine the copy of push notifications for our members, and introduced a chat interface called Blinkist AI that curates hyper-relevant recommendations based on user queries
In addition to finding more content that they’ll enjoy, our users are often looking for new content types that might not match their past behavior. Here, an interesting challenge is providing users with the right mix of recommendations – an “exploration vs. exploitation” issue.
For this, we often must look both at short-term (e.g. whether users interacted with the content we recommended to them) and long-term behavior (e.g. do users come back to consume more content days or weeks later?). Therefore, we are always iterating on our machine learning models to provide users with the optimal personalized content discovery experience. As a result, we empower learners with tailored experiences that enhance motivation, engagement, retention, and satisfaction.
We are just at the beginning of our AI journey at Blinkist, yet we can already see that the potential of AI for learning and development is awe-inspiring. Some benefits are:
As we journey forward, expect a flurry of exciting innovations propelled by AI, including:
As the market leader for bite-sized learning, we recognize that the rise of Generative AI has led to concerns about its potential to replace Blinkist. This concern, however, will only become a serious threat if we choose to remain passive and let our competitors overtake us.
Rather, embracing this technological shift proactively puts us in an ideal position to reap its benefits. With our extensive user base and vast experience in content production and learning, we possess a powerful advantage. These assets require years of cultivation and are the perfect foundation for us to effectively utilize Generative AI on a large scale in our industry. As you see, AI does not signal the end, but rather the start of a transformative journey that allows Blinkist to enhance learning and continue being pioneers.
I hope that you are as curious and enthusiastic about AI as we are. And I invite you to join us on this journey of learning and discovery and embrace the future with us.
