Following the development of go Siasun Robot&Automation Co.Ltd(300024) alphago and artificial intelligence prediction protein folding system alphafold, Google's AI company deepmind recently announced the creation of heavy programming Siasun Robot&Automation Co.Ltd(300024) system alphacode, and the programming level has reached the average level of human programmers.
Deepmind said alphacode's ability to write computer programs has reached a very competitive level. In the competition with human programmers, alphacode can reach a medium level, ranking among the top 54%, marking an important step towards independent coding.
Oriol vinyas, deepmind's chief research scientist, said the study was still in its early stages, but preliminary results showed the possibility of creating an autonomous system to address the challenges of coding programs. "At present, only human beings can do it in this field." Vinyals said, "in the long run, we are excited about alphacode's potential to help programmers and non programmers write code, improve productivity or create new ways of making software."
Alphacode is still unable to beat the best human programmers. Critics also warned that machine generated code must be examined particularly carefully for errors or unforeseen effects, including vulnerability to cyber attacks.
But deepmind ultimately hopes that this AI tool can solve open-ended problems that require critical thinking, logic and language understanding. Vinyals said that in the short term, alphacode can become an assistant to programmers, while in the long term, the system will be able to write its own software one day.
Mike mirzayanov, founder of codeforces, a website that challenges alphacode, said in a joint statement with deepmind: "I can say with certainty that alphacode's results have exceeded my expectations. I think alphacode has reached the level of a promising new competitor. But implementing the algorithm is not the most difficult. Inventing the algorithm is the best way."
Alphacode is not the first AI tool to generate computer code. Many technology companies are developing similar applications. Last June, Microsoft and AI Lab openai adapted the language generation program into an automatic program that can complete the code string. For end users, these systems are like Gmail's intelligent writing function, which can complete any content that people are writing.
"Although human programmers may still have advantages today, AI will surpass humans in many fields. It's only a matter of time." Xiao Jianxiong, founder and chairman of autox, an automatic driving company, told the first financial reporter.
AI is considered to be the most effective tool to help human beings understand the world around them, and it is also the battleground of technology giants. Deepmind was acquired by Google at a price of US $600 million in 2014. Microsoft funded openai with us $1 billion in 2019. Meta also has a strong AI research department. These laboratories lead the world's most advanced AI research and development.
Deepmind co-founder and CEO demis hassabis once said: "deepmind's ultimate vision has always been to build a general AI and accelerate the pace of scientific discovery and research to the greatest extent."
However, it is still difficult to rely on AI to make profits. Deepmind made its first profit in 2020 after losing £ 500 million in 2019, with a profit of about £ 44 million, but the source of profit was not disclosed. Last year, the company released the alphafold algorithm for predicting protein folding, which has entered the field of drug research and development.
A founder of a Chinese artificial intelligence software enterprise told the first financial reporter that deepmind's income source is limited and the investment is huge. Deepmind does not sell products directly to consumers, nor has it announced any deal with private companies other than its parent company, alphabet. However, it does sell software and services to alphabet's departments, including Google, youtube and moon landing plan X.
But most experts believe that the tasks that can be realized by the human brain can be realized by a physical system in theory. "It's just a matter of time and money." Computer expert Herbert jaegar told China business news, "see who is willing to pay a huge sum to develop these technologies."
Deepmind currently has about 1000 employees. High salary is an important means for deepmind to attract talents. The company employs scientists from the world's leading artificial intelligence research field. Most of these top talents have diplomas from Oxford, Cambridge, Stanford and MIT, and their annual salary usually exceeds one million dollars.
