An outstanding conference that I have recently been to, was CogX, which could be described as ‘everything you need to know about AI”. It is fascinating how quickly new technologies and experts of those technologies are emerging. A couple of years ago, artificial intelligence and robots taking over people’s jobs were more like a science fiction, but we already have a 2-day event with 7 stages and some entertaining showcases — including robots like Sophia or Robothespian, the Roborace cars, a live cyberattack monitor, etc. — around the topic with meaningful content. It was also great to see Carina on stage throughout Day 2, where she was not only the host of the Cutting Edge stage, but she also shared a bit of Episode1’s investment philosophy with the audience.
Sophia, the social humanoid robot
It would be difficult to cover such an extensive event in a blog post (and I only attended Day2), so I will only share some of my impressions here that I found thought provoking. My CogX experience started with one of the biggest dilemmas around AI, notably, ETHICS & HUMANITY.
- Esther Dyson opened the morning with emphasising how transparency around ethics is key when talking about AI. She mentioned autonomous cars as an example. Many of us have probably tried the ‘game’, where we are to decide what a self-driving car should do in certain situations (e.g. save the driver’s life or the pedestrian’s life?). Developers, regulators are facing the same challenges, but it is for real. Therefore, it is essential that these programmed rules will be visible to everyone.
- There are a lot of articles about the dark side of artificial intelligence claiming that AI is evil. On the contrary, we could see some wonderful examples how machines can make people’s life complete. Haiyan Zhang, the Investment Director of Microsoft Research Cambridge, gave an insight into her incredible job. She shared the story of Emma, who has Parkinson’s disease. The Emma Watch that Haiyan created is a great illustration of how technology can be utilized not to perfect its wearer but to enable humans.
VIDEO: Emma’s story
- The speakers from Autodesk raised a very exciting topic with ‘bringing humanity and EQ (emotional intelligence) to AI’. They briefly walked us through the story of AVA (Autodesk Virtual Assistant), and what it took to create a human-like voice and virtual assistant, who customers actually like. I often hear that people find certain robots / virtual assistants rigid or scary. To prevent the development of such disappointing machines, it is important to create a diverse development team and include diversity in datasets that the algorithms are trained on. This way, the likelihood of creating a machine on biases is much lower.
The next big question to deal with is what VALUE AI can provide to BUSINESSES.
- Keynote speaker, Manoj Saxena, gave a very interesting overview of the misconceptions that enterprises have regarding artificial intelligence. One of them is ‘AI takes my job away’, while there is much more opportunity in AI powered job augmentation than thought. Another mistake is to confuse AI with BIG Data and Neuro-Linguistic Programming, Machine Learning & Deep Learning. I really liked the analogue that he used to clarify this misunderstanding: Big Data is like senses that do not automatically make someone intelligent; NLP, ML, DP are only tools that enable pattern recognition. He also mentioned that the implementation of AI is very poor: 99% of startups operate AI as a Black Box, which are not explainable and not compliance ready.
- During the following panel discussion, it was agreed that ‘the best AI is when you do not see AI’. In a corporate environment, AI is a strategic capability to build intelligent processes. It continuously learns from feedback but is not visible to customers. The only thing that a customer should experience is that he/she gets more personalised services.
- If data is labelled properly and has enough depth, algorithms will also add value by increasing the scalability of certain processes (e.g. better detection of fraud and money laundering in financial institutions).
- ‘Responsible AI’ in the enterprise context was also covered on the Blockchain stage. Automation can be very useful to avoid intuition in business decisions (e.g. retail expansion). While a ‘robotic management’ will not be afraid of making those decisions and will not walk out of the door when something goes wrong, it is again a question, who in the organization will take the responsibility for the decisions of a machine?!
One of my favourite topics is MOBILITY & SMART CITIES, so I spent a decent amount of time to listen to experts in the field.
- One of the speakers emphasised that we should stop talking about ‘smart cities’. We should talk about ‘better cities’ instead that are not technoratised and not monopolised by tech companies. This idea takes us to the factors that would be needed for ‘smart cities’ to take off and become ‘better cities’: 1. smart leadership; 2. relationship with the private sector; 3. involvement of citizens.
- Standardisation of mobility data would be essential. Currently, it is a big obstacle for creating smart cities. It also matters where to place the sensors to make sure that the right indicators are measured. Startups that are able to tackle this problem will make fortunes.
- The combination of electric and self-driving cars will bring the era of Taas (Travel as a Service). While today, the cost of shared services (e.g. Uber) is double the cost of transportation by an owned vehicle, it is estimated that in 10+ years, TaaS will have a cost advantage.
TaaS
Source: Frank Chen’s presentation
- Urban planning will be the milking cow for those who master it as the changing means of transportation will require the redesign of current road structures. E.g. It is possible, that less lines will be needed due to the optimized traffic enabled by autonomous cars.
- A funny thought that shows the radical changes that AI based technologies can cause: if there will be less accidents due to self-driving cars, movie makers will need to come up with new drama themes.
CogX also embraces the BLOCKCHAIN technology.
- The most exciting discussion — or we could call it more a debate — was around the role and acceptance of Stablecoin and specifically, Maker. If you have not heard of this concept, do not blame yourself! It is quite new. Here is a good summary to better grasp what stablecoin means and what it brings to the table.
Finally, it is always interesting to know how investors relate to a space. The VC Panel revealed some interesting insights into the DEEP TECH landscape.
- Panellists highlighted that now there are more fundamental enablers than before, meaning that the computational power is increasing thus, the technology advancement has accelerated thus, VC investors are more likely to take the risk that scientific R&D involves. It is also encouraging that deep tech teams can now communicate across disciplines and they pick up salesmanship realising its importance.
Investor Panel discussing Deep Tech
#Deeptech #SeedtoseriesA #Startups
