Artificial intelligence (AI) can now write computer code. Human programmers, however, do not have to worry.
The New York Times assures us that AI writing computer code is “good news for humans” and not a threat to professional programmers. To elaborate, Codex, a new application from OpenAI, can write code for basic programs.
OpenAI Codex instantly passed the coding challenge Alphabet (GOOGL) and Facebook (FB) give to programming job applicants, Tom Smith tells Times writer Cade Metz. Smith claims Codex wrote a program that replaces all the spaces in a sentence with dashes and a program that identifies invalid ZIP codes instantly. OpenAI Codex then completed several other coding tasks fast.
OpenAI is the AI research and development company Elon Musk founded. OpenAI’s mission to build artificial general intelligence, or highly autonomous systems that can outperform humans at economically valuable work. If the claims about Codex are true, it sounds as if OpenAI is achieving that goal.
Code Writing AI is available on GitHub
Additionally, Codex can generate programs in 12 computer languages and translate between them. If you want to see OpenAI Codex in action, go to GitHub Copilot. Copilot autocompletes code snippets for GitHub users, The Verge reports.
OpenAI Codex powers Copilot. Copilot can import tweets and attract GoodReads ratings. Frighteningly, Copilot is already available on Microsoft (MSFT) subsidiary GitHub’s website.
Among other tasks, they claim GitHub Copilot can covert comments to code. Additionally, GitHub copilot can suggest tests for implementation code. It can also suggest coding alternatives to programmers, just as grammar programs suggest alternative words to writers. You can even GitHub Copilot to generate code for social media such as Twitter, GitHub claims.
Yes, Code Writing AI will Kill jobs
“These are problems that would be tough for a lot of humans to solve, myself included, and it would type out the response in two seconds,” Smith tells Metz. Smith runs AI company Gado Images. “It was spooky to watch.”
Frighteningly, others familiar with Codex made similar statements. “It is a way of getting code written without having to write as much code,” Jeremy Howard says of Codex. Howard founded the Fast.ai artificial intelligence lab.
“You can tell it to do something, and it will do it,” programmer Ania Kubow says of Codex.
Similarly, the GitHub website claims “GitHub Copilot works great for quickly producing boilerplate and repetitive code patterns. Feed it a few examples and let it generate the rest!”
So yes, Codex could take humans’ jobs and throw highly paid tech bros onto the unemployment line. Yet, both Smith and Metz have the gall to claim Codex is just a tool that can help programmers work faster.
AI can generate programs in 12 Languages
I think Smith’s own statements disprove that contention. Smith said; and Metz reported that Codex can perform programmers’ jobs.
For example, Metz reports Codex can translate code from one programming language to another. Additionally, OpenAI Codex can generate programs in 12 computer languages.
I’m no expert in programming and AI, but Metz’s article and the statements he collected say Codex can perform some programmers’ jobs. Yet the two are eager to suppress the obvious news.
Is The New York Times lying about job-killing AI?
Smith’s interest in hiding Codex’s job-killing capacity is obvious. Metz claims Smith runs an AI company known as Gado Images. Hence, Smith could save money; and increase his own income by using Codex instead of hiring programmers.
Metz’s motivation is harder to discern. I suspect Metz is trying to adhere to The New York Times’ party line on tech. That party line is that tech does not kill jobs and technological unemployment is a fantasy. Moreover, The Times promotes the obvious fallacy that new tech always creates new jobs.
Yes, tech usually creates jobs, but it also kills jobs. Moreover, they design a lot of tech to kill jobs or replace humans. Banks deploy automatic teller machines (ATMs) to reduce the need for human tellers, for example.
Will OpenAI Codex replace human programmers?
Similarly, I imagine the purpose of OpenAI Codex is to replace human programmers or reduce the need for them. Hence, Codex is a labor-saving device and labor-saving devices often kill jobs.
Yes, Codex could help some programmers perform more work and make money. However, I think it could reduce the need for programmers and eliminate jobs or freelance positions.
Yet Metz, and his editors, must deny that obvious reality. I suspect Times editors fear offending affluent upper-class readers who profit from job-killing tech. For example, investment bankers and investors who own stock in Big Tech companies.
Sorry New York Times OpenAI Codex will kill jobs
Notably, Metz tries to reassure educated Times readers that tech will not take their jobs. Instead, he arrogantly claims tech only threatens the likelihoods of working stiffs in warehouses.
In particular, Metz refuses to ask the obvious question: why did they create OpenAI Codex? What is its purpose?
I suspect the purpose of OpenAI Codex is to save money by eliminating the need for computer coders. Indeed.com estimates that the average US computer programmer or coder’s salary was $48,381 per year in 2021.
That math will sound great to an executive trying to follow a budget. However, it will sound terrible to the computer programmer who finds himself driving Uber to pay the bills. The math will frighten all the people who accumulated mountains of college debt to get “high-paying” tech jobs, only to find that OpenAI Codex and other AI are killing those jobs.
Technological Unemployment is Real
However, the party line at The New York Times is tech does not kill jobs. Thus, staffers must ignore or suppress any facts, or circumstantial evidence, that contradict the party line. Worst of all, that party line now applies to Times news coverage.
I have no problem when opinion columnists such as Ross Douthat, the NYT’s tame conservative, promote the party line. Voicing or writing opinions is Douthat’s job. However, it is the job of reporters, such as Metz, to report the news warts and all, not to spin news to keep readers happy.
Frighteningly, I think The Times’ management adopts that party line because it is the message their readers in America’s leadership classes want to read and hear. Nobody in America’s leadership classes, expect Andrew Yang and Elon Musk, admits technology is killing tens of thousands of jobs and will kill millions of jobs soon.
I think America’s leaders will need to confront technological unemployment soon. Those leaders will have no choice when legions of unemployed programmers and coders politicians promised “high-paying IT jobs” in exchange for massive student loans march on Washington and New York demanding answers.
Unfortunately, our leaders will not pay attention to the obvious reality of technological unemployment until the bricks fly through the windows at The New York Times.