Books, Courses & Certifications
What Will Kids Lose If PBS Gets Cut?
Near a cardboard cutout of Daniel Tiger, a small stuffed version of Curious George and plenty of promotional posters in the PBS Kids office, there sit thick stacks of graduation invitations. Most are accompanied with handwritten letters from students extolling the influence children’s television shows had on their journeys to donning the cap and gown — one fresh grad writes that she plans to become an elementary school teacher thanks to PBS.
Sara DeWitt says that while the office has seen its fair share of letters over her two-plus decades with the network — fielding scores of wedding invitations and even more to birthday parties — it has not received so many graduation announcements until this season.
“The outpouring of support is helping remind us why this work is so important and what an amazing impact it has on lives,” DeWitt, the PBS Kids senior vice president and general manager, says. “We see this outpouring as proof of the thoughtfulness and intentionality of the media we’re creating — and that it works.”
The deluge of encouragement comes amid a flurry of actions from the U.S. Department of Education and the White House moving to pull national funding from the Public Broadcasting Service. Justifying the ordered change, the Trump administration argued that spending public money on media groups like PBS through the Corporation for Public Broadcasting is “not only outdated and unnecessary but corrosive to the appearance of journalistic independence,” especially considering there are “abundant, diverse, and innovative news options” in today’s media landscape.
The funding cuts would threaten to dismantle public television, long seen as a safe viewing space for children and parents alike.
As PBS leaders fight the loss of funding, they argue that it may not only spell the end of PBS programming like “Arthur,” “Clifford the Big Red Dog” and “The Cat in the Hat Knows a Lot About That!”; it could also be detrimental to the foundation of research focused on children’s media.
And when there are more options than ever for children’s entertainment, advocates say that producing research-backed, high-quality, non-commercial options for families — particularly those who live in low-resource areas — has never been so important.
‘No One Else Is Researching as Much’
While most parents trust PBS programming — citing it as more trustworthy than any other media source for 22 straight years — many do not know the guardrails put in place to ensure shows are both informative and entertaining, giving the one-two punch necessary for educating children.
“They don’t necessarily understand production, but they sure are appreciative,” says Shelley Pasnik, principal investigator for Ready to Learn programming, a 30-year effort from the federal government that helps to develop educational media. “Once they start to engage and have the space to slow down, they think, ‘There’s a reason I trust the media coming from PBS Kids.’ It’s joyful, and educational, and we’ve heard that in our formal research process.”
Dave Peth, the creator and executive producer of PBS show “Lyla in the Loop,” has worked on other educational media in his 20-plus years in the industry, and he says “no one” deploys the level of rigorous research and testing used in PBS programming.
“Lyla in the Loop,” for example, goes beyond showcasing a family of six living in a Philadelphia-esque city. Peth originally began developing the show in 2015 to focus on computational thinking, which deploys strategic thinking patterns commonly used in engineering and computer science. Nine years later, the show premiered.
“It’s not uncommon for a PBS broadcast series to take a fairly long time to develop,” he says. “Yes, it does take extra steps to make sure what we’re building is based on solid research on what works in education, but it’s worth it.”
PBS works with advisers — ranging from educational researchers to psychologists — who create a framework of learning goals based on studies and topics that are age-appropriate for children. Producers use those frameworks when creating content for the network — whether it’s a televised show or a game on the PBS Kids app — while ensuring it remains engaging and fun for children. PBS also brings in research evaluators, like Pasnik, who take proposed stories and present them to children, evaluating their comprehension and engagement. Any takeaways and adjustments are made in the final story and applied to future episodes.
There is also a large focus on “child-centered content,” designed specifically for the age of the target audience and how much they can process. For example, most PBS Kids episodes are 11 minutes, accounting for children’s shorter attention spans and how much they can retain in a single sitting.
“PBS allows producers to take the time and do it right; we don’t take shortcuts,” Peth says. “You step back and realize, ‘Yes, we are making a contribution,’ to the media landscape and to kids’ and families’ lives.”
The research is particularly important because, as a public media company, PBS regularly and publicly posts its findings for others to build upon.
Federal funding, like the Ready to Learn grant, accounts for about 15 percent of the PBS total budget, costing each taxpayer roughly $1.40 per year, according to PBS. PBS also receives support from foundations, programming dues — and, as many will recall hearing at the end of each PBS episode, from viewers like you.
The Ready to Learn grant saw its funding from the U.S. Department of Education cut in May, prematurely ending its current five-year run, leaving $23 million untouched and stopping its research work immediately.
If this slashed federal spending leads to programming cuts, proponents of the network say it will be tough to replicate the scale of what PBS produces, along with the decades of research conducted by the corporation and the know-how to deploy it.
“It’s like asking, ‘Don’t you think other universities can do the kind of high-quality research Harvard is doing?’ No, I don’t,” says Kathy Hirsh-Pasek, a director of Temple University’s Infant and Child Laboratory. “They don’t have the people, the labs and the sustained support.”
‘A Benefit to Society’
In addition to the research and programming being done at a national level, PBS is also in the unique position of spanning about 330 local stations. Most of those affiliates work directly in their own communities, offering workshops, camps and other engagement efforts.
“They’re bringing this programming that builds off these characters that children love and relate to, and bringing the learning to them,” says Seeta Pai, vice president of education and children’s media at the PBS Boston-based affiliate GBH. “That’s what these stations are uniquely suited to do; they’re the boots on the ground.”
Those local outreach programs are particularly important in what some call “low-resource” areas, meaning places where children live in lower-income households and have less access to broadband internet or information centers like libraries.
“I see [PBS programming] as a resource for those that may not have access to other material goods,” says Rachel Barr, professor and chair of the department of psychology at Georgetown University. “What’s been found, again and again, is that access to educational content is more predictive for learning, for academic outcomes and social outcomes. And again, the effects are strongest for families that don’t have access to other material resources.”
The studies showcasing the positive effects of PBS on children’s learning seem endless. A 2015 study showed children who watched “Super Why!” had stronger literacy skills. That same year, a study found viewers of “Peg + Cat” had stronger mathematical skills. A 2021 study found “Molly of Denali” had better problem-solving skills. Several researchers that EdSurge interviewed pointed to a study from the University of California, Los Angeles, asking teenagers — the first to have grown up watching “Daniel Tiger” — about the show, with almost all respondents not only remembering it but also specific episodes and lessons learned.
“We hear on social media almost daily about something like that,” Pai says. “There’s the short- term impact with children’s learning but it’s also a benefit to society. Kids who had more early childhood education are likely to do better in school and life; that prevents societal expenses later on down the road, whether it be crime or poverty.”
And with roughly half of U.S. children not attending any formal early childhood education program, the supplement of PBS’ research-backed programming could make a difference for their future academic and social-emotional performance.
For years, PBS supporters have argued that government leaders should consider these stakes before reducing support for public media. Mr. Rogers famously testified to that effect in front of the Senate in 1969:
***
***
Video courtesy of Wikimedia Commons.
More recently, in 2023, an appropriations bill proposed eliminating federal funding for the Corporation for Public Broadcasting. Shortly thereafter, advocacy coalition Protect My Public Media asked local broadcasting stations how they would be affected. Roughly 230 responded, nearly all stating that the loss of federal funding would cause “imminent” cuts to staff and programming. Twenty-six stations confirmed that they would be forced off-air, and 23 more stations would need to reduce their coverage areas.
That threat was eliminated. But now, faced with its current threat, PBS has already started shrinking. GBH, which created PBS standouts like “Arthur” and “Molly in Denali,” laid off some staff earlier this month, while at the national level, PBS furloughed roughly a quarter of its Kids division.
The Paradox of More Media Than Ever
President Trump’s executive order calling for cuts to PBS argues that there is more media than ever to access. Indeed, the YouTube Kids app amassed over 145 million downloads in 2024, and the majority of streaming networks all offer “kids” profiles stuffed with shows like “CoComelon,” “Bluey” and “Ms. Rachel.”
But that embarrassment of riches ironically makes choosing high-quality programs more difficult than ever for families.
“We’re all awash in content possibilities, but much like parents say it’s a full-time job reading emails for children’s schools, it can feel like a job to find content beneficial for kids,” Pasnik says.
Children spend plenty of time on screens regardless of the content, equal to more than two hours of their day on average, according to Common Sense Media, a nonprofit focused on media and its suitability for children. Screen time only increases when accounting for lower- income versus higher-income homes. According to the most recent census from Common Sense Media, children from lower-income households (those earning less than $50,000 a year annually) spend nearly twice as much time with screens compared to those from higher-income households (which make $100,000 or more a year).
Hirsh-Pasek, of Temple University, compares media consumption to a diet: If you cut out nutritious food, children will either turn toward more unhealthy food, like desserts, or eat less in general, akin to going hungry. She views the funding hit against PBS in the same vein.
“It’s creating a digital desert,” Hirsh-Pasek, who also serves as a senior fellow at the Brookings Institution, says.“Our high-quality programs are the nutritious stuff. There is so much out there that isn’t good for kids. If you take away the stuff that is [good], you’re leaving kids with digital junk food.”
Starting in the 1990s, the Children’s Television Act required broadcast television to air a dedicated amount of educational content and limited advertising during children’s programs. The rise of streaming and online entertainment undermines that guardrail.
“The expansion of the media landscape is a little jarring; we are having kids watch TikTok and Instagram Reels and YouTube videos that don’t have a foundation of research,” says Amaya Garcia, director of preK–12 research and practice at the think tank New America. “Just because you can access it on YouTube for free doesn’t mean that content is high-quality and appropriate.”
Many entertainment options for children claim to be educational and have good intentions but still lack the research-backed methods employed at PBS. Baby Einstein, for example, was regarded as quality programming in the mid-1990s, eventually selling to Disney. However, several studies found that it created no additional benefits, with one even finding it inhibited babies’ language development.
“Researchers can see what children attend to — and they may attend to a lot of things, but they may not learn from it,” Barr, the researcher from Georgetown University, says. “That’s where the PBS grants look at what children are gaining, versus attending. And that’s the difference between a business model and an educational model.”
Garcia has seen the media landscape change even among her three children. With her oldest, born in 2008, “We watched lots of PBS,” she says. She did less of that with her second child. With her third, born in 2019, there was a pivot toward watching shows via the PBS app.
“It’s definitely changed and gotten harder as the kids have grown up, but I also had the foundational experience of looking at media, of what is good and bad,” Garcia says. “The bottom line: We want high-quality public media that’s accessible to kids. Even in light of the evolving media landscape, we still need something parents can trust and rely upon.”
GBH’s Pai believes younger parents especially, who have grown up with screens, have less understanding of what makes for high-quality programming.
“As the tsunami of content has increased, there’s also an increased need for media literacy,” she says. “It’s almost like we’re educators making the curriculum in school: There’s a level of expertise that we bring. And the brand equity is so high in terms of trust … but it’s almost like they’re taking it for granted that it’s there.”
Those working on PBS shows or for the PBS corporation were all hesitant to speak about the organization’s fate as the funding fight continues, instead focusing on highlighting the benefits the network can provide for children in the interim.
“I can’t possibly predict what’s going to happen, but what doesn’t change is people’s need for growth, and kids’ need to expand their minds and gain new skills,” says Peth, of “Lyla in the Loop.” “So as long as that very human need exists — producers like me and others, and PBS, are going to continue to make content to serve that need.”
The year 2025 for AI education provides choices catering to learning style, career goal, and budget. The Logicmojo Advanced Data Science & AI Program has emerged as the top one, offering comprehensive training with proven results in placement for those wishing to pursue job-oriented training. It offers the kind of live training, projects, and career support that fellow professionals seek when interested in turning into a high-paying AI position.
On the other hand, for the independent learner seeking prestige credentials, a few other good options might include programs from Stanford, MIT, and DeepLearning.AI. Google and IBM certificates are an inexpensive footing for a beginner, while, at the opposite end of the spectrum, a Carnegie Mellon certificate is considered the ultimate academic credential in AI.
Whatever choice you make in 2025 to further your knowledge in AI will place you at the forefront of technology innovation. AI, expected to generate millions of jobs, has the potential to revolutionize every industry, and so whatever you learn today will be the deciding factor in your career waters for at least the next few decades.
Books, Courses & Certifications
Artificial Intelligence and Machine Learning Bootcamp Powered by Simplilearn
Artificial Intelligence and Machine Learning are noteworthy game-changers in today’s digital world. Technological wonders once limited to science fiction have become science fact, giving us innovations such as self-driving cars, intelligent voice-operated virtual assistants, and computers that learn and grow.
The two fields are making inroads into all areas of our lives, including the workplace, showing up in occupations such as Data Scientist and Digital Marketer. And for all the impressive things that Artificial Intelligence and Machine Learning have accomplished in the last ten years, there’s so much more in store.
Simplilearn wants today’s IT professionals to be better equipped to embrace these new technologies. Hence, it offers Machine Learning Bootcamp, held in conjunction with Caltech’s Center for Technology and Management Education (CTME) and in collaboration with IBM.
The bootcamp covers the relevant points of Artificial Intelligence and Machine Learning, exploring tools and concepts such as Python and TensorFlow. The course optimizes the academic excellence of Caltech and the industry prowess of IBM, creating an unbeatable learning resource that supercharges your skillset and prepares you to navigate the world of AI/ML better.
Why is This a Great Bootcamp?
When you bring together an impressive lineup of Simplilearn, Caltech, and IBM, you expect nothing less than an excellent result. The AI and Machine Learning Bootcamp delivers as promised.
This six-month program deals with vital AI/ML concepts such as Deep Learning, Statistics, and Data Science With Python. Here is a breakdown of the diverse and valuable information the bootcamp offers:
- Orientation. The orientation session prepares you for the rigors of an intense, six-month learning experience, where you dedicate from five to ten hours a week to learning the latest in AI/ML skills and concepts.
- Introduction to Artificial Intelligence. There’s a difference between AI and ML, and here’s where you start to learn this. This offering is a beginner course covering the basics of AI and workflows, Deep Learning, Machine Learning, and other details.
- Python for Data Science. Many data scientists prefer to use the Python programming language when working with AI/ML. This section deals with Python, its libraries, and using a Jupyter-based lab environment to write scripts.
- Applied Data Science with Python. Your exposure to Python continues with this study of Python’s tools and techniques used for Data Analytics.
- Machine Learning. Now we come to the other half of the AI/ML partnership. You will learn all about Machine Learning’s chief techniques and concepts, including heuristic aspects, supervised/unsupervised learning, and developing algorithms.
- Deep Learning with Keras and Tensorflow. This section shows you how to use Keras and TensorFlow frameworks to master Deep Learning models and concepts and prepare Deep Learning algorithms.
- Advanced Deep Learning and Computer Vision. This advanced course takes Deep Learning to a new level. This module covers topics like Computer Vision for OCR and Object Detection, and Computer Vision Basics with Python.
- Capstone project. Finally, it’s time to take what you have learned and implement your new AI/ML skills to solve an industry-relevant issue.
Become the Highest Paid AI Engineer!
With Our Trending AI Engineer Master ProgramKnow More
The course also offers students a series of electives:
- Statistics Essentials for Data Science. Statistics are a vital part of Data Science, and this elective teaches you how to make data-driven predictions via statistical inference.
- NLP and Speech Recognition. This elective covers speech-to-text conversion, text-to-speech conversion, automated speech recognition, voice-assistance devices, and much more.
- Reinforcement Learning. Learn how to solve reinforcement learning problems by applying different algorithms and strategies like TensorFlow and Python.
- Caltech Artificial Intelligence and Machine Learning Bootcamp Masterclass. These masterclasses are conducted by qualified Caltech and IBM instructors.
This AI and ML Bootcamp gives students a bounty of AI/ML-related benefits like:
- Campus immersion, which includes an exclusive visit to Caltech’s robotics lab.
- A program completion certificate from Caltech CTME.
- A Caltech CTME Circle membership.
- The chance to earn up to 22 CEUs courtesy of Caltech CTME.
- An online convocation by the Caltech CTME Program Director.
- A physical certificate from Caltech CTME if you request one.
- Access to hackathons and Ask Me Anything sessions from IBM.
- More than 25 hands-on projects and integrated labs across industry verticals.
- A Level Up session by Andrew McAfee, Principal Research Scientist at MIT.
- Access to Simplilearn’s Career Service, which will help you get noticed by today’s top hiring companies.
- Industry-certified certificates for IBM courses.
- Industry masterclasses delivered by IBM.
- Hackathons from IBM.
- Ask Me Anything (AMA) sessions held with the IBM leadership.
And these are the skills the course covers, all essential tools for working with today’s AI and ML projects:
- Statistics
- Python
- Supervised Learning
- Unsupervised Learning
- Recommendation Systems
- NLP
- Neural Networks
- GANs
- Deep Learning
- Reinforcement Learning
- Speech Recognition
- Ensemble Learning
- Computer Vision
About Caltech CTME
Located in California, Caltech is a world-famous, highly respected science and engineering institution featuring some of today’s brightest scientific and technological minds. Contributions from Caltech alumni have earned worldwide acclaim, including over three dozen Nobel prizes. Caltech CTME instructors offer this quality of learning to our students by holding bootcamp master classes.
About IBM
IBM was founded in 1911 and has earned a reputation as the top IT industry leader and master of IT innovation.
How to Thrive in the Brave New World of AI and ML
Machine Learning and Artificial Intelligence have enormous potential to change our world for the better, but the fields need people of skill and vision to help lead the way. Somehow, there must be a balance between technological advancement and how it impacts people (quality of life, carbon footprint, job losses due to automation, etc.).
The AI and Machine Learning Bootcamp helps teach and train students, equipping them to assume a role of leadership in the new world that AI and ML offer.
Developers are doing incredible things with AI. Tools like Copilot, ChatGPT, and Claude have rapidly become indispensable for developers, offering unprecedented speed and efficiency in tasks like writing code, debugging tricky behavior, generating tests, and exploring unfamiliar libraries and frameworks. When it works, it’s effective, and it feels incredibly satisfying.
But if you’ve spent any real time coding with AI, you’ve probably hit a point where things stall. You keep refining your prompt and adjusting your approach, but the model keeps generating the same kind of answer, just phrased a little differently each time, and returning slight variations on the same incomplete solution. It feels close, but it’s not getting there. And worse, it’s not clear how to get back on track.
That moment is familiar to a lot of people trying to apply AI in real work. It’s what my recent talk at O’Reilly’s AI Codecon event was all about.
Over the last two years, while working on the latest edition of Head First C#, I’ve been developing a new kind of learning path, one that helps developers get better at both coding and using AI. I call it Sens-AI, and it came out of something I kept seeing:
There’s a learning gap with AI that’s creating real challenges for people who are still building their development skills.
My recent O’Reilly Radar article “Bridging the AI Learning Gap” looked at what happens when developers try to learn AI and coding at the same time. It’s not just a tooling problem—it’s a thinking problem. A lot of developers are figuring things out by trial and error, and it became clear to me that they needed a better way to move from improvising to actually solving problems.
From Vibe Coding to Problem Solving
Ask developers how they use AI, and many will describe a kind of improvisational prompting strategy: Give the model a task, see what it returns, and nudge it toward something better. It can be an effective approach because it’s fast, fluid, and almost effortless when it works.
That pattern is common enough to have a name: vibe coding. It’s a great starting point, and it works because it draws on real prompt engineering fundamentals—iterating, reacting to output, and refining based on feedback. But when something breaks, the code doesn’t behave as expected, or the AI keeps rehashing the same unhelpful answers, it’s not always clear what to try next. That’s when vibe coding starts to fall apart.
Senior developers tend to pick up AI more quickly than junior ones, but that’s not a hard-and-fast rule. I’ve seen brand-new developers pick it up quickly, and I’ve seen experienced ones get stuck. The difference is in what they do next. The people who succeed with AI tend to stop and rethink: They figure out what’s going wrong, step back to look at the problem, and reframe their prompt to give the model something better to work with.
The Sens-AI Framework
As I started working more closely with developers who were using AI tools to try to find ways to help them ramp up more easily, I paid attention to where they were getting stuck, and I started noticing that the pattern of an AI rehashing the same “almost there” suggestions kept coming up in training sessions and real projects. I saw it happen in my own work too. At first it felt like a weird quirk in the model’s behavior, but over time I realized it was a signal: The AI had used up the context I’d given it. The signal tells us that we need a better understanding of the problem, so we can give the model the information it’s missing. That realization was a turning point. Once I started paying attention to those breakdown moments, I began to see the same root cause across many developers’ experiences: not a flaw in the tools but a lack of framing, context, or understanding that the AI couldn’t supply on its own.
Over time—and after a lot of testing, iteration, and feedback from developers—I distilled the core of the Sens-AI learning path into five specific habits. They came directly from watching where learners got stuck, what kinds of questions they asked, and what helped them move forward. These habits form a framework that’s the intellectual foundation behind how Head First C# teaches developers to work with AI:
- Context: Paying attention to what information you supply to the model, trying to figure out what else it needs to know, and supplying it clearly. This includes code, comments, structure, intent, and anything else that helps the model understand what you’re trying to do.
- Research: Actively using AI and external sources to deepen your own understanding of the problem. This means running examples, consulting documentation, and checking references to verify what’s really going on.
- Problem framing: Using the information you’ve gathered to define the problem more clearly so the model can respond more usefully. This involves digging deeper into the problem you’re trying to solve, recognizing what the AI still needs to know about it, and shaping your prompt to steer it in a more productive direction—and going back to do more research when you realize that it needs more context.
- Refining: Iterating your prompts deliberately. This isn’t about random tweaks; it’s about making targeted changes based on what the model got right and what it missed, and using those results to guide the next step.
- Critical thinking: Judging the quality of AI output rather than just simply accepting it. Does the suggestion make sense? Is it correct, relevant, plausible? This habit is especially important because it helps developers avoid the trap of trusting confident-sounding answers that don’t actually work.
These habits let developers get more out of AI while keeping control over the direction of their work.
From Stuck to Solved: Getting Better Results from AI
I’ve watched a lot of developers use tools like Copilot and ChatGPT—during training sessions, in hands-on exercises, and when they’ve asked me directly for help. What stood out to me was how often they assumed the AI had done a bad job. In reality, the prompt just didn’t include the information the model needed to solve the problem. No one had shown them how to supply the right context. That’s what the five Sens-AI habits are designed to address: not by handing developers a checklist but by helping them build a mental model for how to work with AI more effectively.
In my AI Codecon talk, I shared a story about my colleague Luis, a very experienced developer with over three decades of coding experience. He’s a seasoned engineer and an advanced AI user who builds content for training other developers, works with large language models directly, uses sophisticated prompting techniques, and has built AI-based analysis tools.
Luis was building a desktop wrapper for a React app using Tauri, a Rust-based toolkit. He pulled in both Copilot and ChatGPT, cross-checking output, exploring alternatives, and trying different approaches. But the code still wasn’t working.
Each AI suggestion seemed to fix part of the problem but break another part. The model kept offering slightly different versions of the same incomplete solution, never quite resolving the issue. For a while, he vibe-coded through it, adjusting the prompt and trying again to see if a small nudge would help, but the answers kept circling the same spot. Eventually, he realized the AI had run out of context and changed his approach. He stepped back, did some focused research to better understand what the AI was trying (and failing) to do, and applied the same habits I emphasize in the Sens-AI framework.
That shift changed the outcome. Once he understood the pattern the AI was trying to use, he could guide it. He reframed his prompt, added more context, and finally started getting suggestions that worked. The suggestions only started working once Luis gave the model the missing pieces it needed to make sense of the problem.
Applying the Sens-AI Framework: A Real-World Example
Before I developed the Sens-AI framework, I ran into a problem that later became a textbook case for it. I was curious whether COBOL, a decades-old language developed for mainframes that I had never used before but wanted to learn more about, could handle the basic mechanics of an interactive game. So I did some experimental vibe coding to build a simple terminal app that would let the user move an asterisk around the screen using the W/A/S/D keys. It was a weird little side project—I just wanted to see if I could make COBOL do something it was never really meant for, and learn something about it along the way.
The initial AI-generated code compiled and ran just fine, and at first I made some progress. I was able to get it to clear the screen, draw the asterisk in the right place, handle raw keyboard input that didn’t require the user to press Enter, and get past some initial bugs that caused a lot of flickering.
But once I hit a more subtle bug—where ANSI escape codes like ";10H" were printing literally instead of controlling the cursor—ChatGPT got stuck. I’d describe the problem, and it would generate a slightly different version of the same answer each time. One suggestion used different variable names. Another changed the order of operations. A few attempted to reformat the STRING statement. But none of them addressed the root cause.
The pattern was always the same: slight code rewrites that looked plausible but didn’t actually change the behavior. That’s what a rehash loop looks like. The AI wasn’t giving me worse answers—it was just circling, stuck on the same conceptual idea. So I did what many developers do: I assumed the AI just couldn’t answer my question and moved on to another problem.
At the time, I didn’t recognize the rehash loop for what it was. I assumed ChatGPT just didn’t know the answer and gave up. But revisiting the project after developing the Sens-AI framework, I saw the whole exchange in a new light. The rehash loop was a signal that the AI needed more context. It got stuck because I hadn’t told it what it needed to know.
When I started working on the framework, I remembered this old failure and thought it’d be a perfect test case. Now I had a set of steps that I could follow:
- First, I recognized that the AI had run out of context. The model wasn’t failing randomly—it was repeating itself because it didn’t understand what I was asking it to do.
- Next, I did some targeted research. I brushed up on ANSI escape codes and started reading the AI’s earlier explanations more carefully. That’s when I noticed a detail I’d skimmed past the first time while vibe coding: When I went back through the AI explanation of the code that it generated, I saw that the
PIC ZZCOBOL syntax defines a numeric-edited field. I suspected that could potentially cause it to introduce leading spaces into strings and wondered if that could break an escape sequence. - Then I reframed the problem. I opened a new chat and explained what I was trying to build, what I was seeing, and what I suspected. I told the AI I’d noticed it was circling the same solution and treated that as a signal that we were missing something fundamental. I also told it that I’d done some research and had three leads I suspected were related: how COBOL displays multiple items in sequence, how terminal escape codes need to be formatted, and how spacing in numeric fields might be corrupting the output. The prompt didn’t provide answers; it just gave some potential research areas for the AI to investigate. That gave it what it needed to find the additional context it needed to break out of the rehash loop.
- Once the model was unstuck, I refined my prompt. I asked follow-up questions to clarify exactly what the output should look like and how to construct the strings more reliably. I wasn’t just looking for a fix—I was guiding the model toward a better approach.
- And most of all, I used critical thinking. I read the answers closely, compared them to what I already knew, and decided what to try based on what actually made sense. The explanation checked out. I implemented the fix, and the program worked.
Once I took the time to understand the problem—and did just enough research to give the AI a few hints about what context it was missing—I was able to write a prompt that broke ChatGPT out of the rehash loop, and it generated code that did exactly what I needed. The generated code for the working COBOL app is available in this GitHub GIST.
Why These Habits Matter for New Developers
I built the Sens-AI learning path in Head First C# around the five habits in the framework. These habits aren’t checklists, scripts, or hard-and-fast rules. They’re ways of thinking that help people use AI more productively—and they don’t require years of experience. I’ve seen new developers pick them up quickly, sometimes faster than seasoned developers who didn’t realize they were stuck in shallow prompting loops.
The key insight into these habits came to me when I was updating the coding exercises in the most recent edition of Head First C#. I test the exercises using AI by pasting the instructions and starter code into tools like ChatGPT and Copilot. If they produce the correct solution, that means I’ve given the model enough information to solve it—which means I’ve given readers enough information too. But if it fails to solve the problem, something’s missing from the exercise instructions.
The process of using AI to test the exercises in the book reminded me of a problem I ran into in the first edition, back in 2007. One exercise kept tripping people up, and after reading a lot of feedback, I realized the problem: I hadn’t given readers all the information they needed to solve it. That helped connect the dots for me. The AI struggles with some coding problems for the same reason the learners were struggling with that exercise—because the context wasn’t there. Writing a good coding exercise and writing a good prompt both depend on understanding what the other side needs to make sense of the problem.
That experience helped me realize that to make developers successful with AI, we need to do more than just teach the basics of prompt engineering. We need to explicitly instill these thinking habits and give developers a way to build them alongside their core coding skills. If we want developers to succeed, we can’t just tell them to “prompt better.” We need to show them how to think with AI.
Where We Go from Here
If AI really is changing how we write software—and I believe it is—then we need to change how we teach it. We’ve made it easy to give people access to the tools. The harder part is helping them develop the habits and judgment to use them well, especially when things go wrong. That’s not just an education problem; it’s also a design problem, a documentation problem, and a tooling problem. Sens-AI is one answer, but it’s just the beginning. We still need clearer examples and better ways to guide, debug, and refine the model’s output. If we teach developers how to think with AI, we can help them become not just code generators but thoughtful engineers who understand what their code is doing and why it matters.
-
Funding & Business6 days agoKayak and Expedia race to build AI travel agents that turn social posts into itineraries
-
Jobs & Careers6 days ago
Mumbai-based Perplexity Alternative Has 60k+ Users Without Funding
-
Mergers & Acquisitions6 days ago
Donald Trump suggests US government review subsidies to Elon Musk’s companies
-
Funding & Business6 days ago
Rethinking Venture Capital’s Talent Pipeline
-
Jobs & Careers6 days ago
Why Agentic AI Isn’t Pure Hype (And What Skeptics Aren’t Seeing Yet)
-
Funding & Business4 days ago
Sakana AI’s TreeQuest: Deploy multi-model teams that outperform individual LLMs by 30%
-
Funding & Business7 days agoFrom chatbots to collaborators: How AI agents are reshaping enterprise work
-
Jobs & Careers6 days ago
Astrophel Aerospace Raises ₹6.84 Crore to Build Reusable Launch Vehicle
-
Funding & Business4 days ago
HOLY SMOKES! A new, 200% faster DeepSeek R1-0528 variant appears from German lab TNG Technology Consulting GmbH
-
Funding & Business6 days ago
Europe’s Most Ambitious Startups Aren’t Becoming Global; They’re Starting That Way
