1. What are you looking forward to the most when starting at MainlyAI?
If you’re going to build algorithms, which will change the world, you need a swift, smart and fierce crew skilled in navigating stormy seas, capturing whales and dodging sharks, all whilst having great fun at the same time. I feel really lucky to be part of this adventure! Also, the loot is great!
2. What do you think about the pace of AI deployment in the industry and why do you think it’s slow? 😉
When the time comes to invest in your company’s growth, you can choose between spending money on hiring a new domain expert or buying specialized technology to augment the existing workforce. Machine learning and AI promise a technology, which will learn from its environment and its past actions. Building this learning technology is just the first step though. Without training, the AI is neither a domain expert nor specialized enough to be of help to said experts. It will get there eventually. However, some patience and supervision are required.
This is where it becomes tricky. Are you guaranteed to get a decent return on your investment? What if it doesn’t learn well enough? It takes a leap of faith and patience. Most of us, when hungry, prefer to buy the bread directly from the bakery, rather than planting seeds and hoping for a rich crop next year. A hungry mind-set doesn’t plan for the future and will spend all time gathering the lowest hanging fruit forever.
In time, when it becomes more common to see machine learning grow new and better domain experts from humble beginning, the thought of AI investments as a sustainable choice will come naturally. I’m here to help those thoughts along.
3. Where do you see companies could benefit from AI in their work to set and achieve sustainability targets?
There are plenty of areas where we simply can’t write a code that would do the job. As an example, machine learning enables computer vision classification in ways that were virtually unthinkable only a few years ago. You can’t program a car to drive in traffic. However, you can write a machine learning algorithm, which learns from its environment, so that it eventually can drive that car. The world if full of complex dynamics, where not even the best of domain experts can figure out the solutions, which connect our reality. Often when you want to reach holistic goals like sustainability targets, those exact complex dynamics must be sorted out and controlled. That provides an excellent opportunity for employing AI solutions which can learn to see new connections in the data that humans wouldn’t be able to discover otherwise.
4. Companies are not operating in vacuum; many processes are similar. What would be the natural next step in development of AI?
We send our children to school to learn from each other and from past experiences. Lots of work is being done in creating such schools for AI-algorithms, where we can share data and learn from each other in a never ending information brokering. I think these new “botnets” are fascinating constructs which could help mitigate the cost and risk of training new algorithms on your own.
5. Five quick questions with Kristofer:
- XBOX or PS?
- If I play a game I use a real rig and not some puny console.
- Fortran or Python?
- Spotify or Vinyl?
- Quest or First-Person Shooter?
- Tea or coffee?
6. Tell us a fun fact about yourself
- I’ve been playing roleplaying games as a hobby since I was 10.
- I think Mark Lawrence is the best author in the world and I simply can’t be bothered by anything less.
- I think the future should be more like Pokemon rather than Cyberpunk.
- I wrote the first botnet ever using an IRC-server and C code.
- I also wrote first ever Swedish weather reporting public web application.
- I once recommended the use of Beagle SQL over MySQL 2.0, because the future of MySQL seemed pale.