The appearance of the AI pair programmer - The Morning Context

source

8 highlights

• Earlier this month, Alphabet-owned artificial intelligence company DeepMind announced that it has created a system called AlphaCode that can code just as well as an average human programmer.

• Google, for example, is all in on AutoML that helps automate the tasks of applying machine learning to real-world problems. The high degree of automation in AutoML aims to allow non-experts to make use of machine learning models and techniques without any expertise in the area.

• Of course, making its development tools better is key for Microsoft. In 2018, Microsoft acquired GitHub for $7.5 billion. The following year, Microsoft invested $1 billion in OpenAI, a research lab in San Francisco backed by the likes of Elon Musk and Y Combinator’s Sam Altman.

• Copilot uses a new technology called Codex, which is a derivative of Open AI’s GPT-3 language-prediction model.

• Will automated development platforms “replace” programmers? Well, no. At least not in the short or mid term.

• One, the tools suggest only small code snippets to improve developer efficiency. Second, it does not always offer perfect suggestions (even though, like all AI products, it is “getting smarter all the time”).

• There are also prescient concerns about copyright infringement and licensing ramifications. While public data sets are a fair game when it comes to training AI models, using them to build commercial products might be a slippery slope.

• Set up as a separate business in 2015 as a moonshot venture to improve healthcare via data and analytics, IBM is selling Watson Health to a private equity firm.