Among the main industries impacted, we'll look at advertising, e-commerce and entertainment. The benefits of generative AI are considerable, including increased productivity, improved decision-making, and the ability to create innovative solutions. However, this technology also raises ethical issues relating to privacy, intellectual property and AI liability, safety and regulation that need to be taken into account. The aim is to provide an overview of the subject, and to help investors understand the current and future issues surrounding this disruptive technology.

Part 1 - The challenges of generative AI

Generative AI is quite impressive already, with  text chatbots and several tools are already capable of creating images, music and videos. The best-known tools are ChatGPT, Dall-E, Midjourney, AlphaCode, Bard and Synthesia.

Source: Sonya Huang, Sequoia Capital

The large (LLM) language models of these chatbots are based on information already available online. This information is then rendered in such a way that the result withstands a probability test including all natural language codes. 

Generative AI is expected to dramatically improve business productivity, make workers' daily lives easier, personalize the user experience, speed up the transition and synthesis of information, increase idea generation and creativity, and reduce language barriers.

Significant progress is expected over the next few years in :

  • Task automation: Cost reduction and productivity gains thanks to the automation of repetitive processes and tasks. For example, Airbnb CEO Brian Chesky reported that his developers were 30% more productive thanks to the use of GitHub Copilot tools, an AI-powered programmer developed by OpenAI.
  • Innovation: Accelerating research and development, fostering new discoveries and applications.
  • Content creation: Production of personalized and adaptive content for the creative industries (advertising, video games, film, music, graphic design).
  • Data analysis: Improved decision-making through predictive analysis and trend identification.
  • Education: Personalized learning and easier access to education thanks to intelligent teaching tools.
  • Security: Enhanced IT security, but also increased risk of cyber-attacks and manipulation.

According to a recent study by Bloomberg Intelligence, generative AI will become a $1304 billion market by 2032 , representing an annualized growth rate of 42% over the next 10 years.

Source : Bloomberg Intelligence

Source: Bloomberg Intelligence

Part 2 - The big winners

Cybersecurity

Cybersecurity, already a focus of attention, is benefiting from these advances. Even if it becomes easier for AI to detect anomalies, the risk of hacking will surely increase with the democratization of its technologies. Solving them, and all the expertise deployed by companies specializing in cybersecurity, will be all the more strategic.

Cloud computing

Cloud computing is one of the major beneficiaries, given the computing power required. LLMs imply increased consumption of cloud storage. The major public cloud providers (AWS, Azure and Google Cloud) already offer tools and services to support the use of AI models. To take the case of Microsoft, the company invested $1 billion in OpenAI in 2019. It continued its support by participating in a funding round in 2021, where OpenAI sought $100 million. In January 2023, Microsoft made a further investment in OpenAI worth $10 billion. Microsoft now holds a 49% stake in the company, but gets 75% of the profits until the company recoups the money invested. Azure is OpenAI's exclusive cloud provider. Microsoft thus has a great chance of becoming part of tomorrow's world alongside the other cloud players.

Semiconductors

We all know the old stock market adage: in the midst of a gold rush, rather than miners clinging to the dream, it's the pick and shovel salesmen who makes a fortune. Semiconductors already equip all the electronic devices we use every day. The growing demand for computing power will enable chip manufacturers to benefit from this fundamental trend. Nvidia seems best placed to capture a large part of this value creation, thanks to its dominant position in the GPU segment.

Specialized software

Each sector will see its winners emerge one by one. The case of Adobe in graphic design, photo and video editing tools is striking. The latest advances in AI will enable Adobe to offer ever more relevant solutions and attack new markets. Recently, Adobe launched Firefly, a new family of generative artificial intelligence models. This tool aims to take on image generators such as Midjourney or Dall-E. Integrated with Photoshop, Illustrator, Adobe Express and Adobe Experience Manager, it will enable the creation of text effects and the generation of images or videos from text queries. The tool will also be available via APIs for integration into workflows. This is just one example of the application of AI, and Adobe has all the resources it needs to successfully turn the corner. In addition, the company's artificial intelligence (AI) and machine learning framework, Adobe Sensei, helps power Adobe's entire suite of cloud-based software products. It helps users of the group's software products to significantly improve their efficiency by automating and providing intelligent suggestions for many aspects of their work.

Part 3 - Entire business sectors set to be revolutionized

E-commerce

AI represents a great opportunity to produce personalized search results in online marketplaces. Ecomm AI models can learn from previous purchases and browsing behavior to enrich future searches with highly personalized results for each shopper. Beyond learning from previous purchases, LLM models enable search platforms to better understand the intent behind a query by synthesizing related words, tone, style and other nuanced linguistic structures. The combination of easier search processes and more relevant ad results should improve conversion and advertising returns. E-commerce companies such as Amazon, Etsy, Pinduoduo, JD.com, Rakuten or Ebay are sure to benefit. 

Travel planning

Artificial intelligence could transform travel planning by learning from previous experiences and suggesting unexpected destinations, itineraries, accommodations and activities. Key players such as Booking Holdings and Expedia have developed AI interfaces dedicated to consumers, including chatbots based on ChatGPT for travel planning. These chatbots enable efficiency gains by handling unstructured travel requests, providing concise results. Interactive AI tools could significantly improve the relevance of suggestions, and thus increase conversion on the main platform.

Advertising

Digital advertising platforms will be the main beneficiaries of generative AI. Major networks like Facebook or Instagram (part of the Meta Platforms group) are already testing generative AI tools with some advertisers, different backgrounds using test prompts. These tools will be particularly useful for small businesses that can't afford to hire a large in-house marketing team or pay an advertising agency. Today, companies often spend more than 20% of their marketing budget on ad creation and staff. By streamlining the creative process, marketers could allocate a larger percentage of their marketing budget to paid media and showcasing their campaigns on the internet, which would be a considerable advantage for companies such as Meta Platforms, Alphabet, The Trade Desk, Integral Ad Science, PubMatic, etc. 

The video game industry

Generative AI has the potential to significantly influence the design, construction and gameplay of video games. It enables designers to create immersive environments and spend more time on compelling stories, puzzles and game mechanics. AI can generate landscapes, terrain and architecture, and create dynamic content, such as realistic NPCs interacting with players. By integrating generative AI into their workflow, designers will be able to deliver more immersive, realistic games that adapt in real time to player interactions, with fewer predefined scenarios.

The video game industry's costs could also finally take a turn for the better. Whereas the best games cost around $10 million 20 years ago, the cost of these has exploded to $200 or even $300 million. Every item in the cost structure can be optimized using generative AI or machine learning. In addition to easing budgetary pressures, AI should speed up the time-to-market for core games and live services.

Music and text-to-speech

AI models are already impressive when it comes to generating music and imitating human voices. In music, generative AI is likely to become an increasingly invaluable tool for songwriters and composers, creating new compositions that can serve as inspiration or encourage musicians to approach their creative process in new ways. We're also likely to see it used to create adaptive soundtracks in real time - for example, in video games or even to accompany live footage of real-world events such as sport. AI speech synthesis will also improve, bringing computer-generated voices closer to the levels of expression, inflection and emotion conveyed by a human voice. This will open up new possibilities for real-time translation, audio dubbing and automated real-time voice-overs and narrations.

Part 4 - Limits

Content generated by generative AI feeds on sources that are, more often than not, unquoted. What share of the revenue goes to whom? Who is responsible for what is said? What about intellectual property? Let's take an example.

Almost all of us listen to music on Spotify. The Swedish music streaming giant has two problems: on the one hand, tens of thousands of AI-generated songs are proliferating on its site, taking a chunk out of the copyright treasure trove. On the other hand, the tens of thousands of fake listeners are in fact (AI) robots who will inflate the number of listens to... AI-generated music! Artificial streaming, then, which turns the whole royalty system on its head. And these two problems create others: the platform is clogged, the servers are under strain, and as a result, so are advertisers, labels and artists. Already in 2021, music platforms in France contained between 1% and 3% of streams considered fake. And the situation has become much worse: start-up Beatdapp, which specializes in identifying streaming fraud, estimates that at least 10% of streams are fraudulent. Spotify has therefore taken matters into its own hands, removing from its platform several tens of thousands of songs by music start-up Boomy, which in fact account for just 7% of the 14 million songs created by this Californian company. The stakes are high: Spotify has paid 30 billion in royalties since its creation, including 5 billion in 2020, and 7 billion in 2021. This example hints at future battles over copyright in content creation, or future problems over remuneration of creators on platforms.

Here's where AI chatbots get their knowledge :

Source : Statista

The zero marginal cost economy

As Jeremy Rifkin announced in his book The New Zero Marginal Cost Society in 2014, everyone will be able to become a creator, a contributing producer. User-generated content platforms will see an increase in the number of content creators and creations as it becomes easier to manifest one's imagination through these tools. As with everything else, this should not encroach on the work of the best: the best games and the best artists will probably always be a notch above.

Conclusion

The advent of AI in our daily lives will revolutionize the way we create and interact with the world. Most business sectors will be affected, positively or negatively. The democratization of access to these intelligent tools will lower the barriers to entry for newcomers in many fields, so today's winners won't necessarily be tomorrow's winners. Companies that don't have an in-house data infrastructure ready to accommodate these AI models risk being a step behind. Similarly, companies with little raw data will be at a disadvantage.

As an investor, it's important to identify underlying trends such as AI, which is not just a fad. However, this nascent craze can lead to exaggerated valuations, of which we must be wary. Taking the time to see the winners emerge, and buying them at reasonable valuations, seems a wise move. You can check out our selections of stocks that should benefit from the AI trend through this link.