New AI tools, led by ChatGPT, will revolutionize the creation of texts, images, videos or music. His ability to write code made less of an impact, but it was just as impressive. As in other disciplines, there have also been shocking articles in recent months written by experts with the title “The End of Programming” or “ChatGPT Will Replace Programmers in Ten Years”. But the key lies in the details: what exactly are they capable of programming?
First, it is able to solve challenges that deserve a lot of impact. magazine nature This Wednesday published an article by researchers from DeepMind (Google) in which they accelerated the efficiency of a C++ algorithm that had been in use for decades by 70%: “We show how artificial intelligence can bypass the current state of the art by discovering hitherto unknown procedures,” the article says. “What’s really interesting is that the system learned to write a series of instructions without anyone telling it how,” explains Armando Solar Lizama, a researcher at the Massachusetts Institute of Technology (USA) who had access to the article before it was published.
“They told him what instructions he could use and he started experimenting with combinations. The only thing they told him was whether they were good or bad. Only from this was the system able to deduce which instructions he should write and call to create the program, ”adds the expert on AlphaDev, the program that Created by DeepMind.
These advances prove that artificial intelligence (AI) overcomes complex challenges. But how far can you go without human intervention? article nature He explains that they posed the problem as a game for the machine, which was specially trained, and humans also directed the process. It’s important, but it’s not an absolute revolution. “People are so excited,” Solar Lizama continues. “With these AI models, we’ve seen that they can suddenly do things that were previously considered very difficult. But one of the questions we have now is to see what they can do well and what they can’t do. Based on that, how can we reimagine programming to make it More effective? There are already tools that are starting to help programmers, but only as an assistant.
It’s like an intelligent learner
The word “assistant” or “assistant” is often repeated in conversations with programmers and computer educators, as if the AI were an intelligent learner. Although the news is shocking, few expect an earthquake in commerce. As Brigitte Pintka, a professor at McGill University in Montreal (Canada) and co-author of a paper showing that ChatGPT will get A grades in an introductory programming course, explains: “ChatGPT can be surprisingly and deceptively good at creating short programs. But it’s not very reliable. Right now. I would say that in the future it will be important to be able to evaluate and validate automatically generated software to ensure a level of security and reliability.” Therefore, programmers will continue to be essential, perhaps with a different role, according to Pientka: “Saying that we no longer need computer scientists thanks to ChatGPT is like saying that thanks to calculators and Excel, we don’t need mathematicians or statisticians.”
“Saying that we no longer need computer scientists thanks to ChatGPT is like saying that thanks to calculators, we don’t need mathematicians”
Brigitte Pintka, McGill University
The same specialized magazine that published “The End of Programming” in January published an article this weekend titled “Artificial Intelligence Doesn’t Help Programmers”, written by the famous professor Bertrand Meyer. As a programmer, Meyer writes, you can use an assistant to keep you in line and alert you when you’re moving away. He continues, “But that’s not what I get [de una IA]. I get the equivalent of a cocky graduate who is smart and well-read, polite and quick to apologise, but hopelessly, careless and unreliable. This supposed help does me little good.”
In developer forums such as Hacker news There are dozens of open discussions about how and to what extent ChatGPT and other, more specific applications can help. The variety of responses is enormous, but the desperate enthusiasm of a few months ago has waned. Solar-Lezama puts it in the context of another phase in the history of programming: “Programming will change as it has indeed changed in the past 20 years. It was easy to reuse code someone wrote. Today I can write very complex programs in Python in 10 minutes that would have taken months for a large team a generation ago.”
As society becomes more digital, it is critical to improve the code that supports computing in the world.
today in @natureIntroducing AlphaDev, an artificial intelligence platform that uses reinforcement learning to discover optimized computer science algorithms.
How it works? 🧵 https://t.co/2ukGaPey5w pic.twitter.com/dSedrv7eik
—Google DeepMind (DeepMind) June 7, 2023
It doesn’t save much time
Artificial intelligence has become the assistant that allows you to somewhat reduce the time spent on programming work. But not in extraordinary amounts: “At first I was surprised to find that ChatGPT offers good code,” says Daniel Lemire, a professor at the University of Quebec (Canada), who does advanced programming work. “copilot [una herramienta de Microsoft] He can write entire sections of code, as if he can read my mind. But sometimes he is wrong. It’s hard to measure the time I could save maybe 10%. Writing code is a small part of programming, which includes design, testing, standards, discussions, and specifications.
“It will change programming because it has really changed in the last 20 years”
Armando Solar Lizama, Massachusetts Institute of Technology
For these reasons, a programmer job at the moment seems like a reasonable bet for the future. Despite the progress in software automation so far, the employment of programmers is increasing. The only sure thing is that in the future there will be more code, thanks in large part to the flexibility that AI allows, but no less than programmers. “I expect that we will continue to hire more people, regardless of developments in AI,” says Lemire. “The jobs will be different, even completely new, but there will be no decline anytime soon because of artificial intelligence. Students who aspire to a career in the software industry need to continue the path,” he adds.
The efficiency of programmers, not their replacement, will be the great advantage of this revolution, according to Stephen Piccolo, professor at Brigham Young University (USA): “Some companies will certainly hire fewer, at least initially, while the implications are understood. Technology. But in the near future, this technology will make programmers more efficient rather than replace them.”
One of the frequent aspirations that AI could allow is natural-language programming: asking a Spanish-language ChatGPT to write the code needed to build an app, website, or graph. Although something can be done, it will also be limited in scope for the time being. “Natural languages are full of mysteries,” says Emily Morgan, UC Davis professor of linguistics and co-author of a paper on how these systems detect computer errors. “On the contrary, programming languages must be unambiguous. So we want our programming languages to be unambiguous.”
But you can ask for something in Spanish and ChatGPT converts it into a code. The problem then will be finding or fixing problems: “We’re moving towards using natural language interfaces (such as ChatGPT) to help build a lot of our code,” says Morgan. He explains: “You can ask ChatGPT for things like ‘build a template for a website,’ but it will still generate code or HTML. You’ll need to translate the natural language into an unambiguous programming language.”
You can follow The Country Technology in Facebook And Twitter Or sign up here to receive The weekly newsletter.
Subscribe to continue reading
Read without limits