Transcription
The World Economic Forum published a finding in 2026 that should make anyone with a student loan put down whatever they are doing. AI skills carry a 23% wage premium in today's labor market. A bachelor's degree carries 8%. A skill that you can learn in 3 months is now worth almost three times more than 4 years and $80,000.
And here's where it gets interesting. The CEO of Anthropic, the company building Claude, one of the most advanced AI systems in the world, warned the US government in 2025 that AI could wipe out 50% of all entry-level white-collar jobs within 5 years. That same week, his company posted contractor roles paying 150 to 250 dollars an hour. And the people getting those roles are not coders.
Now, one more thing before we get into this. The most hyped AI career of 2023 is already dead as a standalone job. And the skill that replaced it pays three times more. I'm going to show you both. Now, let's get into it.
But before we get into it, quick note on the sourcing. All salary data in this video come from Glassdoor, ZipRecruiter, an April 2026 the International Association of Privacy Professionals 2025 survey of 16,000 professionals, PwC's analysis of nearly 1 billion global job ads, and peer-reviewed research from Oxford Internet Institute. What these roles pay in your specific market with your specific background will vary. I'm giving you what the data shows at scale. Your outcome depends on your effort and your situation. Now, that's the honest version.
Let's talk about what the data actually says. PwC analyzed close to 1 billion job advertisements globally in 2025. What they found were workers with AI skills earn an average of 56% wage premium over equivalent workers without those skills. 56%. Lightcast independently analyzed 1.3 billion job postings and confirmed an average of 18,000 annual salary increase for any role requiring AI skills. And it's any role, not just AI specific titles. Any job where AI was listed as a requirement.
And Oxford University researchers analyzed 11 million job postings and found something that should concern anyone building a traditional career path right now. From 2018 to 2023, demand for AI skills grew 21%. In the same period, mentions of university education requirements in those same AI role postings fell 15%. The market is explicitly moving away from degrees and towards demonstrated skills. The Oxford researchers put it plainly. The traditional path of university education leading to a higher pay is no longer the default for AI professionals.
But there's a second part to the story. While AI skill roles are exploding upward, the entry-level jobs that most people used to start their careers are quietly collapsing. Burning Glass Institute data shows that between 2018 and 2024, entry-level software development roles as a share of total postings dropped from 43% to 28%. Entry-level data analysis roles dropped from 35% to 22%. Entry-level postings overall are down approximately 35% since January 2023. The floor of the traditional career ladder is contracting at the same time that the ceiling for AI skilled workers is expanding. The people who benefit from this shift are not the ones who wait to see how it plays out. They are the ones who understand what skills the market is paying for right now and move toward them.
And this is what the next 15 minutes is about. Quick question before I get into the specific paths. What field are you coming from right now? Finance, healthcare, law, education, marketing, operations, tech? Drop it in the comments below because the paths that I'm about to show you are not one-size-fits-all. And your existing expertise is worth more in the AI economy than most people recognize. Your background is your competitive advantage. Let me show you why.
The first cluster that I want to show you is what I call the no-code entry points. These are the paths where the barrier to start is the lowest and where most people underestimate how high the ceiling actually goes. Let me explain what I mean.
Path number one, AI trainer and RLHF specialist. RLHF stands for reinforcement learning from human feedback. And it's how AI models like ChatGPT, Claude, and Gemini were trained to behave helpfully. Humans compare pairs of AI outputs, rate quality, flag safety, and write better examples. The AI learns from those human judgments. At the entry level, general annotation work. This pays about 15 dollars to 25 dollars an hour. And that is not the interesting part. Here is the interesting part. The same platforms paying 15 dollars an hour for general annotation, Scale AI, Outlier AI, Merkle are paying medical professionals 130 to 450 an hour for the exact same work. Not different work. The same platform, the same type of task. The difference is domain expertise.
A radiologist working as a Merkle domain evaluator earns more per hour from AI labs than they earn in their clinical setting. An investment banking analyst described working as a Merkle finance evaluator as an, I'm quoting this directly, "Hard to imagine a better hourly job from a pay perspective." Merkle's average hourly rate across all their domain experts is 81 dollars per hour. Their senior domain experts with 4 plus years of professional experience earn over 200 dollars per hour. One freelancer documented earning 60,000 dollars in 3 months working AI trainer contracts at 80 dollars an hour starting in late 2025. The AI economy is not just looking for coders. It is looking for people who know things. And the thing you already know from your career, from your education, from your years of professional expertise is exactly what AI labs cannot build without you. For full-time positions, data annotation and equivalent roles average 115,000 per year. Senior RLHF specialists at Anthropic and OpenAI earn 120,000 to 180,000 plus equity. The data annotator role has grown 150% year over year as of 2026. The labeling and evaluation market is projected to reach 17.1 billion by 2030 at a 28.4% annual growth rate. This is not a temporary gig economy play. This is structural. Growing demand.
Path number two, AI quality analyst and LLM evaluator. This role barely existed before 2025. Skillsets 2026 salary guide calls it, and I'm quoting, "A new role that barely existed before 2025." Companies employing large language models need humans to systematically test them, find the failures, document the biases, verify the outputs in real use cases. National average according to ZipRecruiter's March 2026 data, 65,000 dollars per year. AI quality and evaluation specialists range from 80,000 to 110,000. Entry barrier is low. Time to qualify, two to four months of structured self-study. This is the specific point for people building toward path one's domain expert premium. You start as a generalist evaluator. You build domain-specific credibility and the hourly rate follows.
Now, I want to tell you about Allison Harbin. She has a PhD in art history from Rutgers. Not computer science. Not data science. Art history. In October 2023, she contracted with Google Gemini's team to help train their AI system. She was evaluating outputs, providing expert human judgment on quality, accuracy, and nuance. She is now earning over 100,000 dollars a year at a healthcare technology company. She is working remote, no relocation, and no computer science background. CNBC covered her story because it challenges everything people think they know about who gets hired in AI. Her credential wasn't a degree in the right field. Her credential was that she knew how to think with rigor, evaluate complex outputs critically, and communicate precisely. Those are the skills that AI labs need more of, not less of, as the models get more powerful.
And then, there's this. OpenAI, the company that builds ChatGPT, runs a program called OpenAI residency. And it pays approximately 220,000 dollars per year, which is 18,000 per month. Here is who it is built for, directly from their program description. People with physics, mathematics, neuroscience, or software engineering who, quote, lack formal machine learning experience. OpenAI, the company most associated with the AI revolution, paying $220,000 a year for people who don't have AI credentials because what they actually need is people who are exceptionally good at thinking in their respective domains. Allison Harbinson's story is not an exception. The OpenAI residency is not an exception. They are evidence of a structural shift in who the AI economy considers valuable.
I promised you at the start of this video that I'll tell you about the most hyped AI career of 2023 and what actually happened to it. Prompt engineering. In 2013, headlines were screaming about prompt engineer salaries of 175,000, 300,000, no experience needed, the whole cycle. And then in 2025, an HR consultant at West Monroe was asked whether his clients were investing in prompt engineering upskilling. His answer, "Now, it's nowhere." Gartner, the technology research firm, officially declared in July 2025 that context engineering had replaced prompt engineering and they project by 2028 context engineering features will be embedded in 80% of all software tools for AI applications.
Here's what that actually means. Prompt engineering as a stand-alone job title evaporated in about 18 months, but the underlying skill, knowing how to give precise structured instructions to AI systems, got absorbed in every other role. Developers need it, product managers need it, data analysts need it, governance specialists need it. Unique job postings requiring AI skills went from 55 in January 2021 to nearly 10,000 by May 2025. That is a 180 times increase in 4 years. The message is not don't learn prompt engineering. The message is don't call yourself a prompt engineer. Call yourself what the market pays for and I'm about to show you exactly what that is.
The second cluster is for professionals who are already 5, 10, 15 years in their career who have domain expertise and leadership experience and don't want to throw it away. These paths don't require you to start over. They require you to redirect.
Path three, AI product manager. AIPMs don't build the models. They define what the models should do. They manage the road map. They translate between the technical team and the business. They own the product decisions. The US average salary for an AI product manager, $133,000 per year according to the most recent data. Entry range, 100,000 to 120,000. Senior range, 150,000 to 200,000 and above. And Netflix has posted AIPM roles with ranges from 300,000 to 900,000 in total compensation. As of late 2023, there were nearly 6,900 open AIPM postings in the United States. That number has grown substantially since. Industries actively hiring, fintech, healthcare, enterprise SaaS, and every generative AI startup that needs someone to own the product. The no degree pathway is direct. Prior product management, operations, marketing, or project management experience is the foundation. Layer on Microsoft AI product manager professional certificate, 3 months, and you have the credential layer. Most AIPM postings explicitly list experience and portfolio over education. The degree line is becoming optional.
Path four, conversational AI designer. Someone has to design how AI systems talk to people, how the chatbot responds when a customer is frustrated, how the voice assistant handles ambiguity, how the AI guide maintains tone across thousands of different interactions. This is not a coding role. It is a UX, writing, and linguistics role. Backgrounds valued, user experience design, technical writing, linguistics, communication, and English literature. National average according to ZipRecruiter's April 2026 data, $83,000 per year. Senior range from CDI data, 110,000 to 140,000. In time to qualify, 4 to 8 months of structured learning, no computer science degree required explicitly. If you are a writer, a UX designer, a communication professional, or anyone whose career involve getting words exactly right, this field is importing your skills.
Path five, AI safety red teamer. Red teaming in AI means your job is to break the AI before bad actors do. So, find the jailbreaks, expose the biases, surface the confident wrong answers that could cause harm in a medical setting, a legal context, and a financial decision. This role comes from security culture, but it does not require hacking skills. And many of the best AI red teamers come from journalism, psychology, law, and philosophy because the skill is not technical exploitation. It is adversarial creative thinking. LinkedIn posted their own AI safety red teaming role at $55 to $65 per hour. Mid-level full-time range, 120,000 to 160,000. Senior roles at Google, Mandiant, Intopic, and Netflix, 132,000 to 226,000 and above. The smarter AI gets, the more red teaming it needs. As models move into medicine, law, and finance, the stakes of a confident wrong answer rise. The demand for human adversarial oversight doesn't decrease as AI improves, it increases.
Quick check-in before I get to the path that most people have never heard of, of the five paths that I just covered, AI trainer, LLM evaluator, AIPM, conversational designer, red teamer, which one fits closest to where you are right now? Drop the number in the comments, one through five, and stay with me because path six is the one that has 81% Fortune 500 demand growth, a 273,000 senior median salary, and almost zero YouTube coverage.
Path six, and I'm going to tell you up front, this is the role with the highest salary ceiling on this entire list. It is also a role that almost no career YouTube channel is covering and I think I know why because it doesn't have a flashy name and it doesn't sound exciting. AI governance analyst. Here's the context. The EU AI Act came fully in force in 2025. It is the most comprehensive AI regulation in the world and companies deploying what the regulation calls high-risk AI systems in hiring, in credit scoring, in healthcare, in law enforcement, face mandatory compliance requirements. The penalty for non-compliance, fines up to 35 million euros or 7% of the global annual turnover, whichever is higher. Companies need people who can audit AI systems, document decision processes, and assess bias and fairness, and defend their AI deployments to regulators. That is not a coding job. That is law, policy, risk management, and an operations job.
Drops February 2026 Talent Intelligence Report found that demand for AI governance skills at Fortune 500 companies grew 81% year over year. year. Accenture, one of the largest professional services company in the world, posted a responsible AI vendor governance analyst role earlier this year. Their own job posting describes this function as, and I'm quoting directly, "Built from scratch." This role category barely existed before 2024. And here's what the IAPP, the International Association of Privacy Professionals, found in their 2025 survey of 1,600 professionals across 60 countries. Junior median salary for AI governance specialist, $130,000 per year. Mid-level median, $158,000 a year. Senior median, $273,000 a year. That places AI governance specialist at the senior level in the top 3% of all US earners. For technical AI governance roles in the tech sector specifically, median salary is $221,000. For legal and compliance AI governance roles, $205,000 median.
And the no degree pathway is direct. A background in law, policy, human resources, risk management, compliance, or social sciences is the on-ramp. The IAPP AIGP, AI governance professionals, is the industry's leading certification and $649 for members, approximately 8 to 12 weeks of preparation. Exam through Pearson VUE. That is the credentialing layer on top of expertise you may already have. AI compliance professionals who add AI risk expertise to their existing credentials are reporting salary increases of 30 to 40% because the demand is structural, non-discretionary, and every company deploying AI in a regulated context needs this function. And the supply of qualified people is nowhere close to meeting that demand. This is the wide-open door. 81% demand growth, 273,000 senior median, almost no competition from people who know it exists. That is the definition of a competitive advantage.
Before I close, let me give you the fastest documented certification path, because most content on AI certifications either skips them entirely or sends you to boot camps that cost 5,000 to 15,000, and this costs $198. The Microsoft AI-900 Azure AI Fundamentals exam is $99, and no coding required. 2 to 3 weeks to complete, and qualifies for $65,000 entry roles. The Microsoft AB-730 AI Business Professional exam is $99, also no coding, in 2 to 3 weeks, and targets the 85,000 to 120,000 salary band. Specifically proves prompting and workflow skills. Total, $198, 6 weeks, no prior technical background required, and this is the fastest return on investment certification stack under $500 for entering AI product management and AI governance entry-level positions.
For the AI governance path specifically, add the IAPP the AI GP certification, that is $649, 8 to 12 weeks, and that combination, the Microsoft stack plus the AI GP, directly maps to junior 130,000 median salary tier that we just discussed. Under $850 total, under 6 months, a credential stack that maps to a $130,000 role. Apply the $100 rule that I've used before. For every dollar that you spend, does a salary data suggest at least a $1,000 in annual senior premium? Well, with $800 spent, a $130,000 entry salary in a field with 81% demand growth, the math works.
In our Heights community, I've heard from several people who made exactly this kind of move. And one of them spent 12 years as a compliance officer in financial services. When her firm started deploying AI into their underwriting process, she was one of the first people who asked to oversee the audit of it. Not because she knew AI, because she knew compliance. She spent 6 months getting the AI GP certification and building the specific AI risk vocabulary on top of everything she already knew. She didn't change careers, she changed her value proposition within the career that she already had.
If you want the full resource package from this video, the direct links to every certification, the Merkle and Outlier AI application guides, the AI governance path by industry, and a 90-day action plan mapped to your specific background, that's inside the Heights on Patreon. Link is in the description. Now, joining the Heights does not guarantee you a job, an income increase, or any specific outcome. Success depends entirely on your effort, your consistency, and market factors that neither of us control. What I can tell you is that every resource inside is current, sourced, and real.
The degree premium is not what it was. The AI economy does not care where you went to school. It cares what you know, what you can evaluate, what you can govern, and what you can build. The door is open. The only question is whether you're going to walk through it. This is Antoine Wade. Keep climbing, and we'll see you in the next video.