Ingeniero Prompt

Icono
Icono del constructor
Icono
Icono de puzzle
Related roles: Language Model Engineer, AI Dialogue Architect, Natural Language Generation Specialist, Conversational AI Developer, Prompt Design Specialist, Language Model Tuner, Dialogue System Engineer, AI Language Engineer, NLP Engineer (NLP stands for Natural Language Processing), AI Content Designer
El constructor
El solucionador de problemas
Informática y Tecnología de la Información
Focos Hechos Educación Conseguir trabajo Noticias

Focos

Títulos similares

Language Model Engineer, AI Dialogue Architect, Natural Language Generation Specialist, Conversational AI Developer, Prompt Design Specialist, Language Model Tuner, Dialogue System Engineer, AI Language Engineer, NLP Engineer (NLP stands for Natural Language Processing), AI Content Designer

Descripción del puesto

Artificial intelligence has been around since the mid-50s, at least as an academic area of study. The field has exploded in recent decades, with AI now being used in virtually every sector. From “medicine, transportation, robotics, science, education, the military, surveillance, finance and its regulation, agriculture, entertainment, retail, customer service, and manufacturing,” CalTech notes that AI is about to “become ever more influential in our lives.” That’s an understatement! 

The prediction is already coming true with the advent of OpenAI’s ChatGPT-4, Google’s Bard, Microsoft’s new Bing AI, and other competitors in the generative AI race. But how do such AI programs function? How do they “learn” to communicate? In part, through the hard (and patient) work of Prompt Engineers!

There’s a dynamic subfield of AI called natural language processing (or NLP). NLP essentially teaches computers how to learn and understand written and verbal messages in a way that’s similar to how humans learn. It’s up to Prompt Engineers to come up with text-based prompts that help train these incredible AI models. The programs then use another aspect of AI—machine learning (or ML)—to analyze inputs and generate responses. There’s also a subset of ML called deep learning, which trains neural networks to learn by example! 

The point is, as advanced as artificial intelligence is, it’s only that way because of the patient, behind-the-scenes work of Prompt Engineers and other NLP and ML experts. So at least for now, AI still needs us! 

Aspectos gratificantes de la carrera profesional
  • Helping to evolve revolutionary AI models 
  • Working in a transformative industry that has the potential to improve lives in many ways
  • Lucrative salaries and currently strong job prospects
Salario medio
$255,000
La primicia
Responsabilidades laborales

Horario de trabajo

Prompt Engineers work full-time, with possible overtime necessary depending on goals and timeframes. Their duties are performed indoors in office settings, but their work may require them to travel to various locations to complete on-site tasks. 

Tareas típicas

  • Design prompting techniques aimed at improving AI model performance
  • Develop and optimize clear, precise prompts, such as user-generated inputs and predefined instructions
  • Use prefix-tuning to optimize continuous prompts 
  • Convert tasks into prompt-based datasets
  • Train language models via prompt learning 
  • Improve large language model (LLM) capacity (such as the ability to answer questions and solve arithmetic problems)
  • Use chain-of-thought prompting to enhance LLMs' reasoning ability
  • Augment LLMs with domain knowledge; improve LLM safety
  • Rigorously analyze outputs for accuracy and relevance; correct ambiguous prompts and re-input prompt variations into the AI model
  • Manually sort through datasets
  • Screen for harmful responses resulting from pulling raw, unfiltered data. Determine which inputs caused unacceptable outputs
  • Work with text-to-image prompts for AI models that generate images

Responsabilidades adicionales

  • Work with cross-functional teams to solve problems
  • Troubleshoot system failure causes 
  • Safeguard against potential prompt injection attacks
  • Maintain thorough documentation of procedures
  • Keep up-to-date on industry advances and trends
  • Help develop a knowledge base of best practices
Habilidades necesarias en el trabajo

Habilidades sociales

  • Analítica
  • Pensamiento crítico
  • Orientación al detalle
  • Disciplinado
  • Independiente
  • Metódico
  • Paciente 
  • Planificación y organización
  • Capacidad de resolución de problemas
  • Buen juicio 
  • Trabajo en equipo
  • Gestión del tiempo 

Competencias técnicas

  • Knowledge of natural language processing, machine learning, deep learning, and large language models
  • Ability to develop and implement clear prompts
  • Proficiency in prompt scripting languages (i.e., Bash, Python, Java)
  • Prompt automation tools (i.e., Ansible, Puppet)
  • Linux operating systems
  • Command-line interfaces
Diferentes tipos de organizaciones
  • Private business enterprises
  • AI research companies 
  • Universidades
Expectativas y sacrificios

Prompt Engineering is a relatively new career field, and companies are currently scrambling to find qualified workers. However, because the job is so hot and currently pays so well, it is expected to soon attract droves of new candidates from computer science backgrounds. Those with the right education and experience foundations may be able to get trained in this emerging field in only a few weeks or months. But as Prompt Engineer Rob Lennon stated to TIME, “These are jobs that probably only 500 people could do right now, so there are these insane salaries. But in six months, 50,000 people will be able to do that job.” 

Another concern about the future of this field is that there’s a lot of uncertainty. Professor Ethan Mollick of UPenn’s Wharton School acknowledged, “It’s not clear that prompt engineering is going to matter long-term because AI programs are getting better at anticipating what users need and generating prompts.” 

In other words, Prompt Engineers may only be needed until AI can take off the training wheels and propel itself forward without human help. So right now, Prompt Engineers are expected to be methodical and patient as they train AI models, but at some point, they may sacrifice their own ability to find work once AI no longer requires them. 

Tendencias actuales

Generative AI was introduced back in the ‘60s, but only came to the forefront of the AI field in recent years, thanks to generative adversarial networks which empowered models to create authentic text, images, audio, 3D models, etc. Recent breakthroughs, such as Generative Pre-trained Transformer (aka GPT) have shocked the public and caused a flurry of excitement…and alarm! 

With AI now firmly in the spotlight and its capabilities coming more into focus, companies are rethinking their traditional ways of doing business. Several career fields are expected to be dramatically impacted by the rise of generative AI, including computer science-related jobs such as coding. AI can, or will soon be able to perform an enormous range of tasks that formerly only human workers could tackle. The jobs that AI doesn’t replace, it will likely at least augment in some form. 

Even Prompt Engineers are not immune to AI’s sweeping transformation of the workforce, because eventually, AI will be able to do the same tasks as the prompt writers. In a sense, AI could put its own creators out of a job. That said, companies and governments around the world are starting to raise an eyebrow at how powerful AI is becoming, with some demanding a halt to certain areas of AI research.

¿Qué tipo de cosas le gustaba hacer a la gente de esta carrera cuando era más joven...

Prompt Engineers were probably keen on technology at an early age. They may have been interested in computer coding, tinkering with programming languages, or even hacking. At the same time, they could have enjoyed analytical problem-solving, reading advanced books, or writing stories. 

Teamwork is an important part of this career field, but Prompt Engineers must be comfortable working alone and focusing for long periods of time. This ability could have developed through similar experiences in childhood, maybe from programming or gaming for hours on end.  

Educación y formación necesarias

Educación necesaria

  • Prompt Engineers generally need a bachelor’s degree in computer science, computer engineering, or a related major
  • Common course topics include:
    • AI ethics
    • Toma de decisiones 
    • Deep Learning
    • Human-AI Interaction
    • Large Language Models
    • Aprendizaje automático
    • Procesamiento del lenguaje natural
    • Robótica
  • Relevant classes may also cover:
    • Computer systems
    • Differential and integral calculus
    • Functional programming
    • Imperative computation
    • Matrices and linear transformations
    • Probability theory
    • Sequential data structures and algorithms
  • Employers will look for applicants who already have proven experience in NLP, ML, LLMs, and Deep Learning. They may also want candidates with experience in prompt development, prompt scripting languages, Linux operating systems, and prompt automation tools
    • Note, recruiters may care more about one’s knowledge and skills than where those skills came from (e.g., college classes, certifications, bootcamps, or ad hoc, online courses)
  • Many education and training sites are popping up offering lessons in Prompt Engineering, such as Prompt Engineering Institute and Learn Prompting. These may be great for learning beginner, intermediate, and even advanced skills. Just keep in mind, employers might ask to see proof of your knowledge and abilities, so print or save any certificates of completion
  • Linux can be learned through Codecademy and other online sites. Some say it only takes a few days to grasp the basic commands, and a few months to learn advanced commands
  • Students can also learn programming languages like Bash, Python, and Java on their own or through classes
  • Class Central offers details on a range of free Prompt Engineering online courses
  • Also check out offerings from Coursera, such as its ~18-hour Prompt Engineering for ChatGPT course, or DeepLearning.AI’s ChatGPT Prompt Engineering for Developers

We asked ChatGPT for some extra educational recommendations. Here’s what it said: 

“A Master's degree in computational linguistics or NLP would provide a more specialized focus on the language-related aspects of prompt engineering. A Master's degree in machine learning or artificial intelligence would provide a deeper understanding of the statistical and mathematical principles underlying NLG models.” 

ChatGPT also offered the following degree suggestions! 

  • Bachelor's in Computer Science: “Provides a solid foundation in programming languages, algorithms, and data structures that are fundamental to developing effective NLG models.”
  • Bachelor's in Computational Linguistics: “Provides a deeper understanding of the language-related aspects of prompt engineering, including syntax, semantics, and discourse.”
  • Master's in Computational Linguistics: “Focuses specifically on the study of natural language and how it can be processed by computers.”
  • Master's in Natural Language Processing: “Focuses on the development of algorithms and models for processing human language.”
  • Master's in Artificial Intelligence: “Covers a broad range of topics related to intelligent systems, including machine learning, natural language processing, and robotics.”
  • Master's in Data Science: “Provides a strong foundation in statistical modeling, machine learning, and data analysis that are all relevant to prompt engineering.”
  • Master's in Linguistics: “Provides a deep understanding of language structure and usage, which is fundamental to developing effective NLG models.”
  • Master's in Cognitive Science: “Focuses on the study of human cognition and how it can be modeled using computational techniques.”
  • Master's in Human-Computer Interaction: “Focuses on designing and developing user interfaces that are intuitive and easy to use, which is important for prompt engineering applications.”
Lo que hay que buscar en una universidad
  • Decide if you want to complete a degree in computer science, computer engineering, or something related…with a focus on AI!
  • In addition to the major and AI focus, look at a school’s offerings specific to Prompt Engineering
  • Considera el coste de la matrícula, los descuentos y las oportunidades de becas locales (además de la ayuda federal).
  • Think about your schedule and flexibility when deciding whether to enroll in an on-campus, online, or hybrid program. Prompt Engineering is very suitable for online learning! 
Cosas que hacer en el instituto y la universidad
Hoja de ruta típica
Prompt Engineer Roadmap
Cómo conseguir tu primer empleo
  • Try to complete a computer science or related degree, with a focus on AI
  • Get as much practical hands-on experience in NLP, AI, LLMs, etc. as you can before applying
  • Beef up your resume by taking ad hoc courses related to Prompt Engineering, if your college program didn’t offer enough relevant classes
  • Remember, Prompt Engineers should also be proficient in language arts, so showcase any work experiences you may have related to writing or teaching English
  • Check out job portals such as Indeed, Simply Hired, Glassdoor, and Craigslist. Pay close attention to the education and experience requirements listed
  • Check out online Prompt Engineer resume templates and review potential job interview questions
  • Include job-related keywords on your resume so it can get past the Applicant Tracking Systems software. Keywords may include things like: Python, Java, GPT, DALL-E, Midjourney, Bing AI, NLP, knowledge of LLMs, data collection and analysis, machine learning models, image synthesis, research, etc. 
  • Ask working Prompt Engineers for job-seeking tips
  • Talk to your academic advisor, professors and instructors, and the school’s career center for help
  • Let your network know you’re looking for work! The majority of jobs are still found through those you know, so start using some of your social capital! 
  • Before listing personal references on your application, ask those people in advance if they are willing to write letters of reference or to be called by a potential employer

We asked Google’s Bard to offer some extra job-seeking tips. Here’s what it said: 

  • “Learn the basics of large language models (LLMs). LLMs are the foundation of prompt engineering, so it is important to have a good understanding of how they work. There are many resources available online and in libraries that can teach you about LLMs.
  • “Practice creating prompts. The best way to learn how to create effective prompts is to practice. There are many different ways to create prompts, so find a method that works for you and experiment with different techniques.
  • “Build a portfolio of your work. Once you have created some effective prompts, compile them into a portfolio that you can show potential employers. This will give them a good idea of your skills and abilities.
  • “Network with people in the field. Attend conferences, meetups, and other events where you can meet people who are working in prompt engineering. This is a great way to learn more about the field and make connections that could lead to a job.
  • “Be persistent. The job market for prompt engineers is competitive, so don't give up if you don't land your dream job right away. Keep practicing your skills, building your portfolio, and networking, and eventually, you will find the right opportunity.
  • “Focus on your soft skills. In addition to your technical skills, employers are also looking for prompt engineers who are creative, collaborative, and able to think outside the box.
  • “Be willing to learn. The field of prompt engineering is rapidly evolving, so it is important to be willing to learn new things and stay up-to-date on the latest trends.”
Cómo subir la escalera
  • Become an expert on the AI models, software, and programming languages you use
  • Troubleshoot thoroughly and keep refining output
    • Remember, Google's parent company Alphabet “lost $100 billion in market value…after its new chatbot shared inaccurate information in a promotional video and a company event failed to dazzle.” Somebody probably screwed up their chance for a promotion after that one! 
  • Follow and carefully document procedures to ensure your prompting is consistent and accurate
  • Demonstrate that you can be trusted to work independently
  • Ask your supervisor which skills you should improve upon to add more value to the company. If they suggest doing a certification, knock it out (but see if they’ll pay for it)
  • Learn all you can from co-workers with more experience than you. However, don’t take shortcuts or pick up bad habits. Be sure to follow procedures as directed by your employer
  • Collaborate effectively on teams and focus on solving problems
  • Demonstrate leadership by setting an example for others to follow
  • Train new workers patiently and thoroughly. Answer their questions and keep them motivated to learn new things
  • Stay engaged with professional organizations and keep up-to-date on breakthroughs. AI is evolving rapidly and new ground is being broken every day
  • Be aware of concerns regarding the exponential rise of AI, including worries about the technological singularity—a projected future event that would “involve computer programs becoming so advanced that artificial intelligence (AI) transcends human intelligence, potentially erasing the boundary between humanity and computers.” 
Herramientas/recursos recomendados

Páginas web

Libros

Plan B

Prompt Engineering is a hot career right now, but that trend may not last forever. The potential salary range is a magnet attracting tons of new students eager to jump on the gravy train. At the same time, AI continues to advance and could eventually be able to do its own prompting. If that happens, some of the people who rushed to learn prompting skills may find themselves without opportunities to work. We can’t predict the future, but if you’re interested in exploring similar occupations, we suggest the following to consider:

  • Big Data Engineer/Architect
  • Desarrollador de Business Intelligence
  • Computer and Information Research Scientist
  • Director de Informática y Sistemas de Información
  • Ingeniero informático
  • Computer Network Architect
  • Programador informático
  • Especialista en soporte informático
  • Analista de sistemas informáticos
  • Database Administrators and Architect
  • Científico de datos
  • Analista de seguridad de la información
  • Mathematicians and Statistician
  • Ingeniero de aprendizaje automático
  • Ingeniero en robótica
  • Ingeniero de software
  • Arquitecto de software
  • Desarrollador web

Noticias

Ofertas de empleo

Cursos y herramientas en línea