Neural networks promise to double the yield in Kazakhstan, but there is no one to hold the tractor steering wheel in the fields Will artificial intelligence save Kazakhstan's agriculture - panacea or digital cutting? Neural networks promise to double productivity. But 65% of farm machinery is rusty iron with a word of honor and blue electrical tape. A combine manufactured in 1997 will carry out the algorithm's commands, and Uncle Ermak in the Akmola area will wonder why the tractor won't start. Hundreds of billions of tenge are committed to digitalization: Bairak - 270 billion, the Ministry of Agriculture is asking for another 134 billion. And experience suggests: the more zeros there are in the estimate, the greater the chance that a significant portion will go to contractors and consultants who know how to write beautiful technical specifications. At the same time, the country collected 24.3 million tons of grain - the best result in 15 years. These are achievements of real people who plow the land, and do not write instructions. The question is not whether the Kazakh field needs artificial intelligence. necessary. And the question is whether the word "neural network" will become exposed when the budget is cut, and in three years they will be reported as a "platform" that exists only in reports. So the yield will remain the same. And the responsibility is not anyone's.