Programadores de Computadoras
Las personas también vieron
Riesgo de automatización calculado
Alto Riesgo (61-80%): Los trabajos en esta categoría enfrentan una amenaza significativa por la automatización, ya que muchas de sus tareas pueden ser fácilmente automatizadas utilizando tecnologías actuales o de un futuro cercano.
Más información sobre qué es esta puntuación y cómo se calcula está disponible aquí.
Encuesta de usuarios
Nuestros visitantes han votado que es probable que esta ocupación se automatice. Esta evaluación se ve respaldada por el nivel de riesgo de automatización calculado, que estima una posibilidad del 70% de automatización.
¿Cuál crees que es el riesgo de la automatización?
¿Cuál es la probabilidad de que Programadores de Computadoras sea reemplazado por robots o inteligencia artificial en los próximos 20 años?
Sentimiento
El siguiente gráfico se incluye siempre que haya una cantidad sustancial de votos para generar datos significativos. Estas representaciones visuales muestran los resultados de las encuestas de usuarios a lo largo del tiempo, proporcionando una indicación importante de las tendencias de sentimiento.
Sentimiento a lo largo del tiempo (trimestralmente)
Sentimiento a lo largo del tiempo (anualmente)
Crecimiento
Se espera que el número de ofertas de trabajo para 'Computer Programmers' disminuya 9,6% para 2033
Empleo total y estimaciones de vacantes laborales
Las proyecciones actualizadas se deben 09-2025.
Salarios
En 2023, el salario anual mediano para 'Computer Programmers' fue de 99.700 $, o 47 $ por hora.
'Computer Programmers' recibieron un salario 107,4% más alto que el salario medio nacional, que se situó en 48.060 $
Salarios a lo largo del tiempo
Volumen
A partir de 2023, había 120.370 personas empleadas como 'Computer Programmers' dentro de los Estados Unidos.
Esto representa alrededor del 0,08% de la fuerza laboral empleada en todo el país.
Dicho de otra manera, alrededor de 1 de cada 1 mil personas están empleadas como 'Computer Programmers'.
Descripción del trabajo
Cree, modifique y pruebe el código y los scripts que permiten que las aplicaciones informáticas se ejecuten. Trabaje a partir de especificaciones elaboradas por desarrolladores de software y web u otras personas. Puede desarrollar y escribir programas informáticos para almacenar, localizar y recuperar documentos, datos e información específicos.
SOC Code: 15-1251.00
Comentarios
Leave a comment
So its pretty hard to train an ai that can be adapted to every it infrastrcuture. It also is a risk since it would mean giving ai access to 100% of the system, which is a concerning security risk.
One day there may be an ai that can do that, but even then it will require programmers that maintain the ai and check/test code that it wrote since someone will need to take responsbility for what the ai does. And since i cant even gurantee my own code to work at all times in different cases, I sure as hell wont take responbility for some ai code no matter how good the ai is
PR review is getting handled by AI now. Gemini, Copilot, and CodeRabbit are taking over.
Security has always been weak. They just force 2 factor and call it a day. Look at the NPM supply attacks because one guy took the bait on an email.
Please take into account the trajectory of progress rather than the current state of things.
But though AI (= LLMs) has gotten better in the sense of creating more complex outputs, it otherwise suffers from exactly the same problems as early versions: no compositionality, no continual learning, no consistency, and no self-correction.
If you ask it to fix a certain bug fix in a complex codebase, you have a high chance that it also starts to change something completely unrelated. It doesn't really grasp how precisely elements make up the complex whole; instead, it applies pattern-matching, by which it gets misled. You can waste days with AI on what would be a five-minute manual fix.
Also, junior devs' job isn't to churn out mediocre code for generic, long-solved tasks or create the 1000th to-do list app. Instead, they get onboarded and are supposed to familiarize themselves with your codebase to improve it. But you know that "PT" in ChatGPT stands for "pre-trained?" So AI doesn't learn anything new. AI is like suffering from anterograde amnesia.
Anyway, neither apps nor code are like a commodity where "more is better". In the end, the point is to create something new. Otherwise there is not much point: just use an already existing and tested library. And it's exactly the "new" part where AI sucks because it struggles to transcend its training data.
So no, AI is incapable of replacing junior devs.
It would have a lot of immediate positive effects if AI worked like you describe: it would insanely empower open-source projects (often lacking maintainers) to fix all their long lists of bugs or introduce new features. But this is not what we observe at all. Because AI is overhyped and underdelivers.
-> MNCs keep laying off juniors while recording profits. Why? Because they have trained their AIs to do the jobs of 100s of people. Where they needed 100 people, they only need 1 now.
-> OpenAI just hired ex-bankers to develop financial AIs that will eventually replace junior finance professionals in their company. Others are following suit.
-> Salesforce has openly admitted to cutting jobs because of AI. So have many other top companies.
Don't get me wrong, I think AI is way off from completely replacing digital labour. I almost agree with you on most of your points.
But it's getting there, and it's getting there quickly. It's only a matter of time until researchers develop a new system that supersedes transformer models. And judging by the Billions being burned every day on AI research, it's almost inevitable.
I think you'll find the "AI 2027" paper very interesting. Please do have a read. And thanks for your reply :)
Deja una respuesta sobre esta ocupación