Why AGI is THE boost for the Pharma industry

Processes in pharma are always slow and very resource intense – with Artificial General Intelligence (= strong AI) new drugs can be created in an intelligent way!

Normally it needs around 15 years and 1 billion $ to market a pharma product. Picking the compatible molecules is very hard as there are millions of possibilities. So, it is not surprising that it needs up to five years to find the right one. After this phase the animal and later on the human testing has to be done. Only very few drugs survive this process, due to occurring adverse reactions. So, lots of data about unsuccessful trials are available, but cannot be used.

As this slow process cannot keep up with the speed of other technologies only few companies can afford to do research and create new improved drugs. A lot of potential stays hidden in the huge and complex amount of data. Statistical systems are too weak to associate all these data and find new correlations. Also, convenient AI-systems are no able to create new usable knowledge out of the collected material.

In contrary to “normal” AI, which is working with few neural networks and deep learning methods, AGI (Artificial General Intelligenc), is working with thousands of neural networks which are connected with each other and in constant exchange. Therewith for the first-time machine thinking is possible. Machine thinking means that the system is able to create new ideas, find new solutions/associations and react to untrained information properly.

For the pharma sector this means that with AGI suddenly lots of new possibilities to use existing data are occurring.
Especially in three areas AGI has the ability to change the whole market:

1. Drug discovery: Building of new molecules for drugs containing fragments of existing molecules. Pre-testing of the AGI if they can be built in a chemical way. Benefits: Creation of optimal molecules for every disease with less adverse reactions. Massive limitations of chemical testing and animal testing.
2. Usage of mishits: Although a sparse populated database with existing drugs is available, the testing of new areas of application is not possible because of multiple diversity. AGI is able to analyze existing drugs and find new ways of how to compose their molecules to be optimized and reused in new areas. Also, AGI can reduce the number of possibilities massively, so that targeted testing can be possible -> massive saving of time any money!
3. Genetics: Analyzing of the genetic code, recognizing of different diseases and which drugs are having the best effect on certain genotypes/phenotypes. Therewith individualized medication is possible for the first time.

Compared to convenient AI AGI can also work with few or missing data, unstructured data or complete different sources of data. For the pharma areas this means that quantitative and qualitative data can be combined. As AGI can work with all kind of data adaption or programming activities are being completely dropped.

We can´t wait for the first results of the symbiosis of Pharma and AGI!

Isabell Kunst, 11/30/2017