Enjins Scale-ups Nick 18 November 2022

4 Takeaways on AI from the WebSummit 2022

Beginning of November over 70.000 tech enthusiasts gathered in Lisbon for the WebSummit. Next to catching up with current and new clients, the event gives a clear insight into the current state of Artificial Intelligence (AI) within different industries. Here are my takeaways from three exciting days.

1. Expect AI disruption faster in late-adopting industries like legal tech and green tech

AI continues to manifest itself in every single industry. From the early joiners like e-commerce and fin-tech to the relative newcomers like legal tech and green tech. It is more difficult to find an industry where AI does not play a role than the other way around.

As AI has many concepts that are industry-agnostic, it gives the possibility to transfer best practices from high AI mature industries to low AI mature industries. As an example, the legal industry is at the beginning of its AI journey, like e-commerce was ten years ago. However, one of the large differences between those two journeys is that the AI field has figured out quite some innovations. So, for legal tech we do not need to figure out how to deploy AI at scale: we have ML ops best practices. Moreover, we do not need to figure out how to work with both unstructured and structured data. We have found a way through lake houses. Et cetera.

A logical result will be that the potential acceleration of innovation within these industries is even higher; given that constraints are met, like adoption rate, investments, and the availability of high-value use cases. That is why I am really excited about the fields like green tech and legal tech, where the innovative start-ups we have spoken with have the potential to disrupt the market even faster.

2. How to translate AI opportunities or ideas into an actionable plan (both long-term and short-term) remains a challenge

After speaking to many founders, C-level management, and product owners, they have a view of what AI can do, but miss a concrete strategy on how and when to implement AI. Two types of conversations stood out.

  • Firstly, many leaders can come up with 2–3 example use cases. However, the real deal is to fully automate and algorithmicize your organization (both your commercial and your operational model) leading to probably 10+ AI-driven use cases that are linked together. Creating a holistic perspective on AI is a crucial starting point for C-level management to turn Data & AI into your competitive advantage.
  • Secondly, many new platforms, apps, and services remain very functional, rather than intelligent. Meaning, a person can buy an “Article A” for a certain price. That function works well. However, that price is not optimized, nor the recommendations of that article, resulting in a rather unintelligent platform. Especially start-ups that have established a product-market fit, managed to create traffic to their platform and gathered data over time, are in a sweet spot to level up their platform by implementing AI.

3. Generative AI demonstrated its potential in 2022 and will be adopted in 2023

Dall-e dropped a bomb in 2022. Not only within the AI scene but also outside of the AI world. They showed the potential to generate realistic pictures by simply typing in text. Other AI innovation labs followed, by showing use cases to easily generate pictures or videos. As a result, generative AI as a technology has proven itself to be useful.

So, the next question to ask is how the field of generative AI will develop in 2023 and what the impact on different industries will be. Unlike deepfakes — a technology that boomed in 2021 — generative AI has a clearer and larger potential. Simply stated, generative AI will unlock new possibilities for all industries where personalized content generation adds value. Think about social media where users can create even more unique content or the online advertising market where ads are hyper-personalized to the user. Both industries are multi-billion industries. Additionally, applications are found in gaming, fashion, interior design, et cetera. It democratizes design.

Where technological innovation has come from innovation labs like OpenAI (Dall-e) and Google (Stable Diffuser), the mass adoption and acceleration are expected to be led by content platforms (Picsart, etc.) and social media platforms (TikTok, etc.). They determine the success of Generative AI in 2023.

Technology-wise, there are still improvements to be made. Most of the current breakthroughs are around static pictures, whereas most of the current content generated is videos. Moreover, the field of text and audio generation is still waiting for their Dall-e-like breakthrough.

Check out the start-ups below that are working within generative AI. Looking ahead to an exciting year in 2023.

Image by Verve Ventures

4. The barrier to solving complex problems (natural language processing, computer vision) continues to decrease

Many SaaS companies have made it possible for companies to use NLP and Computer Vision, tailored to very specific use cases. Additionally, AWS, Azure, and Google have made these solutions an “API call away”. Think about tools like Amazon Comprehend and Amazon Rekognition. Both developments unlock the possibility for companies to solve complex problems.

How to adopt these solutions then? The best is to do it in an iterative matter:

  1. Check the fit of your use case with already existing API solutions.
  2. Build a benchmark model using an API-ready AI solution.
  3. Tailor & learn to improve the initial “off-the-shelf” solution.
  4. When the solution itself is core to the value of the business, build it fully in-house to reduce dependencies to SaaS and AWS, Azure, and Google.

Looking forward to an exciting 2023 within the AI field. Feel free to share your thoughts in the comments.

Curious to hear your ideas.

Want to stay updated?

Please fill in your e-mail and we update you when we have new content!