July 29, 2024
Generative AI is rapidly transforming industries as it evolves. Read on to learn how generative AI developments are impacting functions, including personalized learning, content creation, and web search, and surfacing the need for responsible AI practices. Reach out to us to discuss this topic.
Generative AI is fast transforming various aspects of the technology landscape. Major updates and launches announced in the OpenAI Spring Update, Google I/O, and Microsoft Build event this year, show how rapidly this technology is evolving. At present, the artificial intelligence (AI) market is marked by technology companies looking to rapidly develop IP and shape eco-systems and standards and by providers and enterprises looking to evolve their business models to absorb generative AI.
Generative AI’s current functionalities and its rapidly evolving capabilities offer much in terms of potential benefits but also come with their fair share of uncertainties. Figure 1 gives an overview of the areas of generative AI impact that we will dive into.
Hyper personalized learning may be upon us
Generative AI promises to personalize learning and make it interactive. It can empower teachers through AI assistants to offer more engaging and accessible learning. Examples of such context-aware AI innovations include OpenAI’s GPT-4o, Google’s LearnLM, and Microsoft’s Khanmigo. GPT-4o offers personalized and adaptive learning, identifying students’ strengths and weaknesses and providing solutions in their preferred learning styles, with multilingual support.
Recent generative AI updates have highlighted advancements in educational tools and platforms, showcasing new features and functionalities designed to enhance personalized learning experiences. Going forward, educators will likely be able to use generative AI tools to customize learning plans for students and understand their learning challenges through data and insights. Perhaps what is even more remarkable is the self-learning potential that generative AI offers. In a world where educators are largely overwhelmed, generative AI may be the force multiplier the education industry has been crying out for.
Generative AI-enabled learning tracks can help organizations thread the needle between scaling L&D initiatives and contextualizing them to different stakeholder needs. Generative AI may have the potential to not only provide 1-1 tutoring on emerging skillsets across a variety of languages but may also be leveraged to ideate and design the curriculum. At a time when the half-life of talent is becoming shorter, generative AI may be the answer to ensuring organizational L&D stays relevant and nimble.
Content creation may soon become commoditized
While generative AI has a wide-ranging impact across the media and entertainment value chain, content generation is where the impact is most acutely felt (see figure 2).
One of generative AI’s most striking use cases has been the creation of hyper-realistic content that seems indistinguishable from artist or studio creations. Recent generative AI updates and advances like GPT-4o have made content generation easier. These technologies can recognize tone, multiple speakers, background noises and produce outputs with embedded emotion such as laughter and songs. Innovations like OpenAI’s Sora or Google’s Veo empower creators and professionals to generate high-quality videos across different cinematic and visual styles without requiring extensive filmmaking expertise.
Advances in content generation have sparked fears about the ongoing relevance and demand for creative roles. Stories like that of Hollywood filmmaker Tyler Perry putting the brakes on a planned US$800 million expansion of his Atlanta studio upon seeing Open AI Sora’s video generation capabilities do little to allay such concerns. While current concerns about AI taking over creative work are understandable, it is more likely that going forward, we will simply see creatives engaged in higher-order work while AI solves for more time-consuming tasks. Content may become more synthetic, i.e., generated as opposed to filmed, produced faster, and more personalized. We may even start tiptoeing towards real-time content generation.
Hyper-realistic generated content also opens the door for deepfakes. False images, videos, and sound clips mimicking public figures or enterprises can lead to public unrest and material damage. With multiple elections being held around the world this year, deepfakes can have a meaningful impact on political discourse. This has understandably led to increased government scrutiny toward generative AI companies. Beyond politics, deepfakes are increasingly being used to commit fraud. Related to this, an employee at a multi-national organization was duped into paying out millions of dollars to those the employee believed to be key stakeholders at the company.
Generated content also poses some interesting intellectual property (IP) questions. Who has rights to the IP of generated content? Is it the person who prompted the output? Is it the technology company whose algorithms are being used to generate the content? Do the individuals or organizations whose data was used to train the algorithm also have some stake in what the model produces? Apart from adherence to IP laws, those using generative AI to create content will also have to be mindful of possible algorithmic biases manifesting in the generated content. Increasing efforts around responsible AI and transparency are needed to ensure biases in training data don’t get reinforced through the usage of generated content to train/tune other models.
Web search is changing
Generative AI is expected to have a massive impact on how web search takes place, and by extension, how online advertising plays out. Consequently, digital advertising, particularly SEO and SEM, are key areas being disrupted by generative AI (see figure 3).
The move from coursing through books at the library to typing out keywords in a search bar was one of the biggest shifts in how people looked for information. Similarly, the impending transition from typing out keywords to simply asking in natural language promises to be the next big shift.
Generative AI updates have introduced new features in search engines and voice assistants, transforming how users interact with these tools. Advertisers increasingly express concerns about bot traffic eating away at their ad dollars. How would they feel about bots being the norm? Imagine if search fundamentally shifts to an audio-visual interface, with those searching for information rarely scanning the website themselves. How might this affect existing advertising models? Here are some possibilities – advertisers may realize that the customer is no longer on the website and needs to be engaged elsewhere. This can lead to a shift in the advertising mix, with more audio ads being rolled into searched information. SERP1 ads may also become more expensive due to their proximity to the search interface. For publishers, the shift may be from using ads to monetize content to directly monetizing the content itself based on how it is consumed to answer questions. Ad exchanges may evolve to become a network for generative AI bots to find content at a given price point. While these are indeed speculations, what is clear is that we are on the brink of a fundamental shift in information search and, by extension, digital marketing. All stakeholders within advertising may have to reassess their role in the broader ecosystem – be it advertisers, publishers, or ad exchanges.
For now, the impact of generative AI on everyday information search is limited. We are starting to see the integration of generative AI tools into existing search engines. For example, Google has integrated generative AI into its search tool. Through this feature, Google can interpret complex visual questions, provide explanations, suggest next steps, and offer resources using an AI overview. Voice assistants like Siri are also getting an overhaul. Apple’s partnership with Open AI promises to provide Siri with generative AI capabilities. The search space had one undisputed king for a long time – generative AI looks to be one of those seismic events that has the potential to reshape this hierarchy.
AI needs to be responsible
Generative AI promises to have a wide-ranging impact across multiple sectors. Given the massive impact that generative AI can have, tech companies need to balance innovation with safety. Responsible AI (RAI) is fast becoming an area of focus for enterprises looking to invest in and scale generative AI. Figure 4 illustrates some key considerations that will shape emerging RAI policies.
Enterprises will increasingly look to collaborate with service providers and technology companies that prioritize data security and have effective governance setups to ensure responsible usage of AI. Implementing ethical guardrails is essential to unlocking the full potential of generative AI and ensuring its responsible usage. As user expectations and government oversight rise alongside AI’s evolution, companies that embrace RAI will be the ones leading the charge in this exciting new era.
If you have questions about this blog or would like to discuss recent developments in the generative AI space, please reach out to Abhishek Sengupta or Oishi Mazumder.
Watch the webinar, Gen AI and the Future of Cybersecurity: Advanced Strategies for Cyber Defense, for insights into new developments, emerging applications, challenges, and opportunities presented by generative AI in cybersecurity.
