There are thousands of multiple companies in Silicon Valley who almost do not contribute value to the daily work of their organizations. They are called “Invisible Programmers”, a famous series has called them “Ghost Engineers.” Unlike what might think, they are not poorly paid engineers, or employees whose performance is not valued. These normal have high wages, but, they really don’t bother “or try.”
And we are not exaggerating but in a survey carried out by the team led by Yegor Denisov-Blanch about a sample of more than 50,000 engineers of the famous American Valley, it was found that almost 10% of these engineers were not bothered or in trying. “
Who developed the study argue several causes of this phenomenon and underline the complexity of measuring the productivity of programmers. Despite the multiple studies to design a master algorithm that supervises the code they publish, the same conclusion is reached: it is not possible to determine exactly the productivity of the developers.
The cause of this is none other than by the nature of the software. Increase the number of lines of code (which might seem the most logical option) does not indicate anything about its quality. In reality, frequently, the quality code tends to be more compact than poor. Additionally, more experience developers dedicate more time to meetings, problem solving and tests, than to programming itself. How then to measure your productivity?
Stack Overflow, after his latest survey he concluded that 68% of the developers consulted did not consider generative artificial intelligence (IAG) as a work threat. Only 12% perceived it as a risk. This percentage is very similar to that of ghost developers (10%). We could think, will they be the same?
Taking into account that in the same survey, 61.8% of developers already use IAG tools, and 13.8% plan to incorporate them shortly, surely yes. According to them, among many other advantages, these tools allow them to be more productive, efficient and accelerate their learning.
There is no doubt that the true threat to a developer is not the AI itself, but another developer that uses it for its benefit. An non -ghost developer, focused on optimizing their performance through tools that improve their development process and increase its productivity. But we return to the same question: how to measure productivity?
We verify that we are approaching the issue that we have been postponing, but before we continue to analyze the impact of AI. The term is already familiar to us
Deepseek. And, if so, we would have to ask ourselves if they have been where have they been in recent months?
The Chinese LLM, Deepseek R1, has broken the rules of the game by drastically reducing the cost of language models training. Thanks to this, access to its API has been offered with a much lower cost compared to those of the main actors in the sector, such as OpenAi.
We can conclude that Deepseek has transformed a weakness into strength. The limitations in the export of microchips have forced to develop an alternative approach to train its model, managing to do so at a much lower cost and with a competitive performance against OpenAI.
Now, seen this, OpenAI quickly reacted partially releasing its O3 Mini model, the best so far, and a much lower price to recover its leadership. However, the damage was already done. Deepseek, despite being a model developed in China, was released as an open source, allowing any company to use it without fear that your data is committed, since it runs on your own machine.
But, in addition, all this allows engineers who are not ghosts, those who seek to improve continuously, have access to advanced tools for the development of code in an accessible and economic way, democratizing their use and multiplying the productivity of developers and the companies that support them.
This idea of software product evaluation is based on ISO/IEC standards and is already being used by governments around the world: the European Union, Japan, South Korea, Mexico, Brazil, Malaysia … just to name a few.
In addition, having this metric, we can identify the “ghosts”, select good developers and grow our business in the right direction.
The software is the protagonist, and its correct evaluation will allow us to optimize not only efficiency but also productivity in software development. It is an objective and verifiable indicator that helps us avoid the subjectivities that often wraps this area.
#Fear #artificial #intelligence #love #Deepseek #invisible #developers
