In the Innovantage podcast, Sigli’s CBDO Max Golikov talks to tech experts and entrepreneurs about their vision of how artificial intelligence is transforming the world. The 4th episode covers much more than that. Leesa Soulodre, who was the podcast guest, explained not only the role of technology in modern society but also the role of society in tech progress.
Check out the full Innovantage episode with Leesa Soulodre here:
Leesa is the founder of R3i Group and the Managing General Partner at R3i Capital, a deep-tech cross-border venture capital firm that focuses on AI and sustainable development.
Leesa’s firm helps projects connect to capital, customers, and non-dilutive financing that lets startup founders keep full ownership of their companies.
Today, founders who work in deep tech (or in other words, who build projects that are based on high-tech innovation or significant scientific advances) traditionally face three main challenges.
While talking about the safety of using AI products, Leesa recollected some other well-known cases. The proliferation of Airbnb made everybody’s homes available to guests. This led to the need for the implementation of trust and safety teams. With the growing popularity of taxi services, everybody’s car can be used as a taxi. This again highlighted the demand for such trust and safety teams.
What do we have to deal with in the case of AI? The situation may look rather alarming. In fact, deep neural network compression technology could be used to kill more people faster and with less energy than other technologies.
And if you consider that every AI product can be used for such a purpose, it becomes obvious that we need to maintain this notion of trust and safety teams. This is vital to make sure that our technologies are not misused for unintended purposes.
Any AI organization with as much influence as OpenAI has that is not going to invest in trust and safety teams will face significant legal and regulatory hurdles. Moreover, such companies will have even more issues with further growth and innovation in the future if they don’t gain implicit trust from their user base.
When it comes to the regulation of tech companies, there always have been some controversies. One of the main reasons for this is that regulations don’t tend to catch up fast enough with what tech companies are doing.
In the discussion of this aspect with Max, Leesa mentioned that she sits on the board of the AI Asia-Pacific Institute. She communicates with representatives of the governments in the region. The governments want to build safety rails for technologies, and AI in particular. But there is a significant barrier.
Let’s take Singapore as an example. The absolute majority of the registered Generative AI companies are just starting their business journeys. They are at their seed or pre-series A stages. This means that quite often they do not even know what they have on the tin. They do not know the value of their products. That’s why it’s natural that they are not ready to invest in regulatory oversight. Given this, there is no sense asking them to do that.
Leesa believes that it will be more sensible to build guard rails into the fabric of the major technologies that underpin new products and solutions created by startups.
For example, many GenAI companies are building their tools on the back of the technologies developed by OpenAI, Microsoft, or Amazon. So it makes more sense to start with these tech giants. They need to comply with regulations first.
While talking about mature AI technologies that startups can rely on, as an example, Leesa mentioned Hugging Face. It is a versatile platform that is widely recognized for its open-source repository of multiple large language models (LLMs).
Leesa’s VC firm works a lot with startups. Around 2,000 projects claimed that they were using Hugging Face. At a close investigation, it was revealed that only 200 were truly using it. And only 16 were fundable in the opinion of R3i’s experts.
Today there are a lot of players with similar offers. Leesa noted that both open-source and commercial models, like ChatGPT, can be a good option for new technologies. But here, it’s vital to understand that they serve different purposes. For example, ChatGPT is perfect for converting high volumes of always the same automation tasks.
As an investor and technologist, Leesa is interested in finding technologies that will work as efficiently and safely as possible and bring the highest value.
She said that she doesn’t invest in ChatGPT-like solutions. She looks for applied AI technologies around critical infrastructure and highly regulated industries. The projects that are the most interesting to her are those that can bring tangible results. They can power transformation from point A to point B in such domains as smart cities, energy, healthcare, industrial manufacturing, water management, agriculture, mobility, space safety, security, surveillance,
For instance, she mentioned that her VC firm often invests in technologies for renewable energy. Already now, it is a highly regulated sector, despite being comparatively new.
In the discussion with Max, Leesa mentioned that today there are a lot of AI-related projects that may look quite appealing for investors. But in reality, they may turn out to be a mouse trap.
Working with the deep tech industry, Leesa prefers to invest only in those projects that have deep scientific research and technological invention behind them, which can be proved by a patent pool or a data mode. And if a project has a data mode, it should be its own, not the one that Microsoft or OpenAI possesses.
But what are these things? And why do they matter?
Leesa explained this with real-life examples. When scientists at the university invent something, they need to protect this from being copied or misused. Almost always in such cases, they can apply for a patent that will protect an idea or an innovation. If somebody else wants to use this innovation, they will need to obtain a license.
However, Leesa warns about one serious challenge related to patents. When you publish a patent, everyone can learn what it is about. Unfortunately, at the moment, patents, especially in the software development industry, are not protected well enough.
Patents themselves can be viewed as assets. Even if a project fails or the development of the technology is frozen, founders will still have a patent that can be further sold.
As for data, it also can be monetized. If your company is carefully collecting, identifying, classifying, and tagging data, you (or somebody else, who will get access to it) can use it to create new products or power the existing ones.
While talking about the capabilities of generative AI, Leesa stated that it is fantastic for ideation, and especially brainstorming. Nevertheless, it’s vital to understand that such models are hallucinogenic. It means that they can provide wrong or irrelevant answers. It may happen because the training data was incomplete or biased. And that is just a bright demonstration of the “Garbage in, garbage out” principle. Moreover, hallucinations may happen because AI models often lack constraints that can limit possible outcomes.
That’s why we can make the following conclusions from such a situation. First of all, we always should be very careful and attentively check whether the received information is true. And secondly, despite the advancements in GenAI, we still need human creativity, empathy, and ingenuity. That’s what AI can’t ensure at the moment.
It’s not a good idea to come to the ball too early, as nobody will turn up at that time. But you also shouldn’t come too late as you will miss all the fun. You should be just in time. That’s why before introducing a new technology, it is necessary to analyze whether the market is ready to receive it as well as to consider the key barriers to its adoption. The psychosocial aspect is important. People should trust you and your solution.
It’s vital to listen to different opinions to detect possible unintended consequences that may stay unnoticed for founders. When a team is working on a new product, they have only one perspective. But when you are building something new for a community, you should know what impact your innovation will have on it. It’s also worth mentioning that the impact on one community may differ from the impact on another one.
It may sound surprising but in many cases, it will be very sensible to listen to children as well. Today, there are even some tech events for kids. And that is a very good trend. One day they will become active users of technology. That’s why their voices, their questions, and their doubts also have value.
Speaking about the technologies that Leesa has invested in, she mentioned a couple of examples. She described them as absolutely revolutionary from the perspective of activation and implementation.
The most promising, value-based technologies should do something at least slightly better than it is done today and can greatly change the way we perceive something. For example, massive carbon emission reduction technologies change the way we think about the use of energy and water.
While talking about new technologies and economic prosperity, Leesa said that one of the biggest concerns is piracy, both physical and digital.
Piracy is one of the factors that can affect supply chains, steal jobs, and put economic development under threat. For example, when digital versions of books are provided for free, people who contributed to their creation lose their wages.
Nevertheless, despite a huge negative effect, sometimes it is possible to detect some positive sides. This way of distribution can play an important role in the digital preservation of some media that no longer generates profit but could still be valuable in terms of history, art, or culture. Moreover, this can open new opportunities for those who have limited access to legal distribution infrastructure.
Different investors may apply their own methodologies to their decision-making process. Leesa shared that at R3i Capital they also have their own philosophy when it comes to choosing projects. One of the most important things that they pay attention to is the team.
To avoid unconscious bias, R3i Capital relies on an AI engine built in cooperation with Hatcher. As a result, every team goes through the same filters and the result of such evaluation is as objective as possible.
Thanks to this approach, absolutely everyone has the same chances. Such a system allows the VC firm to give voice even to those groups that are often ignored, like women and minorities.
Moreover, it’s very important to analyze how fundable the company is and what its likely impact is. According to Leesa, while doing this, it’s highly required to keep 100% transparency. This will ensure the desired trust between founders and capital
After all, capital markets do not need to be brutal. They should be fair instead. This will help to achieve a win-win interaction.
Leesa explained that with its investments, R3i supports tech companies with a tangible ESG product impact. In other words, they focus on products that prioritize environmental issues, social issues, and corporate governance.
However, investors sometimes say that they do not care about the environment and sustainability, they care about money.
But how can a healthcare product not improve patients’ lives if it is a good healthcare product? The same is true about energy, cybersecurity, mobility, and other industries.
Becoming more sustainable, environmentally friendly, and socially valuable doesn’t mean getting less money. In fact, often, it can even mean more money. This can be explained by people’s willingness to pay for the things that can enhance the quality of their living.
If an offered product harms people or has massive negative consequences, society won’t trust it.
As venture capital firms are interested in long-run outcomes, they try to make bets on winning technologies. Sustainability businesses that focus on social and environmental effects are definitely among them.
At the end of this discussion with Max, Leesa shared her thoughts about the role of society in innovations. One of the key recommendations that she can give to everyone is to be more empathetic to each other and common problems.
Sometimes when founders can’t get financial support from governments or corporations, they can receive help from other people who care about solving the problems addressed by their projects.
To achieve success, it’s very important to take the next first step. And we can’t do it alone.
At Sigli, we share this vision and that’s one of the reasons why we create the Innovantage podcast episodes.
If you are also fascinated with the capacities of AI and other emerging technologies, as well as their power to change the world, stay with us. New inspiring ideas are coming soon!
.svg)
.svg)
.svg)
.svg)
.svg)
.svg)
.svg)
.svg)
.svg)
.svg)
.png)
.png)
