Summary
- Artificial intelligence for SMEs – possibilities and limits
- Successfully used in medicine
- Big Data – the problem with AI programming
- Podium with the head of the State Chancellery SH
After the welcome by WFL managing director Ulf Hahn, Kathrin Ostertag took over the moderation of the evening from the IHK Lübeck.
Previously, speaker Stefan Tecuceanu, CEO of MeetPlace GmbH, introduced almost 50 visitors to the subject of artificial intelligence (AI).
The graduate mathematician explained that the AI basically differs from programmed computer technology by a feed back function. This allows the machine to evaluate the success of their actions, identifying problems and “improving” (on the next try).
In the artificial generation of knowledge from experience – “machine learning” – the (artificial) system does not simply act according to its programmed contents, but recognizes patterns and laws and can transfer them.
Dare to wear it in your pocket!
Successfully used in medicine
For example, if a machine is programmed to travel from A to B on a course / direction and speed road, it must fail to be blocked by a job site unless the programmer has entered information. An intelligent aggregate, on the other hand, evaluates the obstacle and changes its behavior “on the next attempt” until it succeeds, ie bypasses the blockage.
Areas where AI has been used successfully include computers for games such as chess or go, game shows or medical diagnostics (e.g., radiology) and robotics. Internet market leader Google used Artificial Intelligence to cut its energy consumption, resulting in an unbelievable 40-50 percent increase in AI. The savings were more than $ 100 million.
“Of course, there are also fears of what Artificial Intelligence is doing,” says Stefan Tecuceanu.
There is the fear that it will cost our work or do something without our knowledge or that we do not want. Tecuceanu: “There will be no jobs left over by the AI, but new ones will emerge.” The fundamental problem in the current situation in the development of AI, Tecuceanu says, is that it requires a tremendous amount of data. In other words, in order to be able to act intelligently, all parameters for every possible situation must be fed into the machine. And the decisive factor here is the high quality of the data.
Conclusions – Dare to wear it in your pocket!
You can find the source on the link WFL event for entrepreneurs in the Möllner Quellenhof