GitHub CEO: Why AI Won't Replace Junior Engineers Anytime Soon
Not according to GitHub CEO Thomas Dohmke.
In a candid conversation with The Pragmatic Engineer, Dohmke dismissed the idea that AI could fully replace junior developers. Instead, he emphasized that early-career engineers are not only still relevant—they're crucial in the AI era.
Dohmke described the notion of AI replacing junior engineers as 'backwards,' arguing that these newcomers offer exactly what the tech world needs: openness, adaptability, and a different perspective. 'Often the folks that are younger in career bring a new perspective to the team and say, 'Hey, why don't we try this?' or 'I want to incubate this idea,'' he said.
He pointed out that GitHub has recently renewed its investment in its internship program, highlighting the value of fresh energy and diverse backgrounds within the organization. 'It's just a really positive thing to hear those folks bring fresh ideas, a great amount of energy and often, you know, a different, diverse background into the company,' he noted.
According to Dohmke, today's younger engineers have grown up with AI, smartphones, and cloud computing as part of their daily lives. This, he believes, gives them a distinct edge when using tools like GitHub Copilot. 'They get to use AI much faster—they get it because they are taking this with an open mind,' he said. 'They haven't been in an experience where some change has led to a big outage. So they're more open-minded.'
He further explained that the expectations for modern developers are shifting. 'The goal of the future engineer is no longer to write it all from scratch,' Dohmke said. 'The goal is to combine their prompting skills and agent open source libraries into getting that problem solved much faster than they could have done two, three years ago.'
The GitHub CEO stressed that both junior and senior engineers are vital to building effective teams. He advocates for hiring based on hands-on contribution, not just tenure. 'You hire people because they have a green contribution graph on their GitHub profile. That matters more to us than whether you have five years at one company and five at another.'
In short, as AI reshapes how code gets written, it's the next generation of engineers—armed with creativity, curiosity, and Copilot—that might just lead the charge.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Hans India
3 days ago
- Hans India
Former Twitter CEO Parag Agrawal Returns with $30M AI Startup 'Parallel' to Challenge GPT-5 in Web Research
Almost three years after being abruptly ousted from Twitter by Elon Musk, Parag Agrawal is making a high-profile comeback in Silicon Valley. This time, the former Twitter CEO is leading his own artificial intelligence venture — and it's already drawing attention for outperforming some of the biggest names in the field. Agrawal's new company, Parallel Web Systems Inc., founded in 2023, operates out of Palo Alto with a 25-person team. Backed by major investors such as Khosla Ventures, First Round Capital, and Index Ventures, Parallel has raised $30 million in funding. According to the company's blog post, its platform is already processing millions of research tasks daily for early adopters, including 'some of the fastest-growing AI companies,' as Agrawal describes them. At its core, Parallel offers agentic AI services that allow AI systems to pull real-time data directly from the public web. The platform doesn't just retrieve information — it verifies, organizes, and even grades the confidence level of its responses. In essence, it gives AI applications a built-in browser with advanced intelligence, enabling more accurate and reliable results. Parallel's technology features eight distinct 'research engines' tailored for different needs. The fastest engine delivers results in under a minute, while its most advanced, Ultra8x, can spend up to 30 minutes digging into highly detailed queries. The company claims Ultra8x has surpassed OpenAI's GPT-5 in independent benchmarks like BrowseComp and DeepResearch Bench by over 10%, making it 'the only AI system to outperform both humans and leading AI models like GPT-5 on the most rigorous benchmarks for deep web research.' The potential applications are wide-ranging. AI coding assistants can use Parallel to pull live snippets from GitHub, retailers can track competitors' product catalogs in real time, and market analysts can have customer reviews compiled into spreadsheets. Developers have access to three APIs, including a low-latency option optimized for chatbots. Agrawal's return to the tech scene comes after a turbulent 2022, when Musk completed his $44 billion acquisition of Twitter and immediately dismissed most of its top executives, including him. That move followed months of legal disputes over the takeover. Rather than taking a break, Agrawal dived back into research and development. He explored ideas ranging from AI healthcare to data-driven automation, but ultimately zeroed in on what he saw as a critical gap in the AI landscape — giving AI agents the ability to reliably locate and interpret information from the internet. Now, Parallel positions him back in the AI race, and perhaps indirectly, in competition with Musk. Agrawal sees the future of AI as one where multiple autonomous agents will work online simultaneously for individual users. 'You'll probably deploy 50 agents on your behalf to be on the internet,' he predicts. 'And that's going to happen soon, like next year,' he told Bloomberg. With speed, accuracy, and reliability as its edge, Parallel could become a defining player in the next phase of AI innovation.
The Hindu
3 days ago
- The Hindu
Islands of excellence amid a morass of stuttering institutions
78 Years of Freedom Pointing to the monkey frolicking outside his room, V. Kamakoti, Director, IIT-Madras, said even the monkey in IIT wants to learn only Machine Learning. His observation may point to where Indian higher education is poised nearly eight decades after Independence. In India today, a bunch of top-level and highly acclaimed institutions produce brilliant and well-equipped minds eager to delve into the latest. But after the so-called Tier 1 institutions, there is a steep fall in quality in Tier 2 and 3 — the vast majority of India's engineering institutions catering to lakhs of students who might be as eager as the IITians to learn cutting-edge concepts but are just not up to the mark. For instance, V. Madhosh Kumar is a rideshare cab driver in Chennai. He says this is a temporary job that will help him find his bearings in Chennai. He has an engineering degree in AI and ML from a college near Coimbatore. Madhosh wanted to know if doing a course on network security would boost his job prospects that appeared nil. 'My professors had little clue and much of what we students learned was self-learning,' he said, adding recruiters don't seem interested in him. Madhosh did have a LinkedIn profile but it indicated little engagement with companies that may be interested to employ him. He said he did upload his college project on GitHub but it had been downloaded only once. Madhosh was clearly not employable. The situation in engineering may apply to other streams as well — a few elite institutions and then an abyss in quality. Poor quality is endemic across India's education — basic and higher. And it's only now that it has caught the attention and getting the focus it has always deserved from policymakers. Some 15 years ago, the Annual Status of Education Reports (ASER) were received with deathly silence, especially on the part of the government. These reports showed that by Class 6, the average student's literacy and math skills are likely that of a Class 3 student. When he is leaving middle school, the average student is likely at the level of Class 6. In 2009-10, students from two of India's most educationally advanced States, Tamil and Himachal Pradesh, participated in the Programme for International Student Assessment conducted by the Organisation for Economic Cooperation and Development. Kyrgyzstan was the only country that ranked lower than India. Across India, postgraduates struggle to write a three-sentence paragraph in any language. Many get cheated on the interest they need to pay on loans they have taken because they just don't know the math. In the book India Is Broken, Ashoka Mody writes about how, unlike in Japan, Jawaharlal Nehru's programme for boosting farm productivity failed. His land reforms weren't implemented properly with the average farmer having little long-term stake in the farms. Further, India's farmers were not literate and aware enough to make use of agriculture extension services. Sub-opitmal delivery As ASER reports kept coming, after a point, governments were forced to take notice. Their schools were just not performing. In a report, NITI Aayog notes that quality deficiency in Indian school education is caused by sub-optimal delivery structures (sub-scale schools, large scale teacher vacancies), and weak organisation structures, governance, and limited accountability (poorly defined organisational structure, ineffective systems, process and accountability). Change has to happen in all three areas. 'For change to scale across the State and to sustain, it is essential to anchor it within the State and not have it led from outside,' the study notes. India Is Broken talks about how India's planners, right from Independence, just did not pay attention to the quality of education delivered by government primary schools even as they were setting up the high performing higher education institutions. And that told on the literacy and math skills of the average Indian. Well into the 1990s, Mody says, as India achieved near-universal enrolment in primary schools, the problem of quality couldn't be wished away. The RTE Act played a role in creating the basic infrastructure, even if the implementation was patchy. And the NEP 2020, for the first time in a government document, acknowledges the severe deficiency in Foundational Literacy and Numeracy, says Balaji Sampath, education activist who has helped to conduct the ASER surveys. Now there is a consensus across India regarding quality of education after the dots have been connected. Poor quality of basic and higher education leads to deficiency in labour productivity — India's labour productivity is 20% of Malaysia's. And poor labour productivity is among the key contributors to why India just doesn't seem to be reaping the demographic dividend that it deserves. Meanwhile, with the IT boom, engineered outside of government planning and intent, the opening up of the economy expanded the service sector spearheaded by IT growth. Now, India hopes that a consequence of this growth will help India go back and complete the circle — high technology as a low-cost solution to the problem of quality. A typically Indian jugaad for fixing a chronic problem. The NEP 2020 does talk about tech and its facilitators such as the concept of self-learning. Mr. Kamakoti of IIT-Madras sees it as a question of reaching the best learning resources to every student in villages and cities through technology. 'It's natural that very accomplished teachers may not be willing to go to villages. The number of teachers available to take up jobs is limited,' he says. He cites the Swayam platform of video classes helmed by IIT professors to emphasize how video can reach quality learning resources over Internet. He talks about interesting translation projects driven by AI that will translate the content to more than 20 languages. 'There are tools to make the sessions more interesting. Animation, virtual reality can help explain subtle, detailed concepts in an intricate manner,' says Mr. Kamakoti who sees the quality problem as essentially the problem of not having enough skilled, high quality teachers as well as learning resources. 'Today, the entire school chemistry lab can be done through Google. All of Class 10 experiments can be done through Google and tests conducted using virtual reality,' he says, adding that scaling up such initiatives is the way to go. Video learning is far superior to conventional classroom learning from just the teacher, says Marmar Mukhopadhyay, education management expert. He recalls an experiment he did in Gujarat where he made a video with still photographs and voiceover on the Gujarat earthquake. He did three types of sessions: in one the teacher played the video from beginning to end; in the second the teacher operated the controls as asked by students; and in the third the videos were given to the students. Learners were asked to write interpretative essays on earthquakes in Japan and the experience of the students' elders with earthquakes. He says the students who were given the videos did best in the assessment and all students scored nearly 80%. Marmar looks at AI as changing the role of the teacher from instructor to facilitator of self-learning, largely. Learning to learn is more important than the learning itself, he says, adding the teacher's job is to create learning opportunities. Choosing the right tech But does tech deliver? Rukmini Banerjii, CEO of Pratham Education Foundation that conducts the ASER surveys, emphasises the importance of field studies and randomised control trials to asses which types of tech can deliver and to what extent. Meanwhile, Sweden had a blowback when the new Conservative government in 2023 sought to roll back digitisation of education programme of the previous government. Swedish neuro-researchers published many papers discussing the negative impact of screentime, fall in socialising among teens, etc. This came after nearly 15 years of digital-first approach to education. 'The reliance on digital tools has led to a lack of critical thinking and overconfidence in online sources,' said Sweden's Minister of Education, Lotta Edholm. The Swedish government has sought to go back to providing printed textbooks to all students for all subjects and in-person classes. Research has indeed shown that students retain more information from print. The sensory experience of flipping through the pages and the absence of distractions does seem to facilitate deep understanding. Yet, while Sweden may be spoiled for choices, India's needs are dire in a resource-starved environment. Several studies by The Abdul Latif Jameel Poverty Action Lab (J-PAL), however, bear out the benefits of use of tech. For instance, one study relying on evidence from 126 randomized evaluations shows that educational software that helps students learn at their own pace deliver. Technology nudges such as message reminders work. Combining in-person and online can work although students in in-person-only courses perform better than those in online-only courses. A targeted, phone-tutoring programme to improve Math by J-PAL South Asia, along with NGOs Youth Impact and Alokit, is being implemented in Karnataka by the government. The programme had been validated in Botswana. The target is to reach more than 4.5 lakh students between grades 3 and 5. For more than a decade now, various States have sought to leverage tech to improve education reach and quality with mixed results. Experts say that AI's ability to personalisze learning could help fix a uniquely Indian problem — teachers are far too burdened with teaching an array of things and concepts, cover the portions, rather, as well as with administrative work. They simply cannot ensure every student has grasped the fundamentals. With student-teacher ratios still a challenge, personal care continues to suffer that AI can execute at low-cost. Mr. Kamakoti notes that the planned Center of Excellence in AI coming under the Ministry of Education can help draw up a nationwide roadmap on AI use. Primary students can submit their worksheets for Language and Math practice every day across the year to an AI System for auto-evaluation to provide teachers and parents with learner data analysis and deliver a personalised practice work that progressively is personalised to the individual learner's skill situation, says Bhanu Potta, EdTech expert focused on social investments and achieving sustainable development goals. Such AI-supported use cases can be built and deployed at large-scale for millions of learners, at a fiscal allocation of Rs. 1,000 per year or less considering the running costs and upfront costs amortised for five years when built on sovereign models, he adds. Viplav Baxi, an education professional with over 30 years of experience in education-technology, says experience has shown that our top-down 'educratic' systems across the world have created similar challenges for teachers. Teachers are our force multiplier and it's time to blend approaches that celebrate and encourage local scale autonomy. 'We must provide them all the necessary resources and tools to help them become more effective and efficient,' he says, adding teachers are indispensable and they provide the last mile delivery.
The Hindu
4 days ago
- The Hindu
Should coding be mandatory learning in schools? Benefits, challenges, and the road ahead
In a world increasingly driven by technology, the question of whether coding should be a mandatory subject in schools has gained prominence. As smart phones become an extension of our hands and AI tools begin to shape how we live and learn, the ability to understand and create with code is becoming more than a technical skill—it's a new form of literacy. This article explores the reasons behind the growing demand for coding in education, the challenges of implementing it, and whether it should truly be made mandatory for all students in Indian schools. What is code? Why is it important? At its core, coding is the process of writing instructions for computers to perform specific tasks. From the apps we use every day to self-driving cars and medical robots, coding is the language behind the machines. Learning to code isn't just about becoming a software engineer. It teaches logical thinking, problem-solving, creativity, and precision and structure. In essence, coding is a skill that enhances 21st-century learning, much like reading and writing did for earlier generations. Shift to teaching coding world Countries like the United States, United Kingdom, Estonia, and Singapore have already integrated coding into school curricula. In the U.K., children start learning programming as early as age five. Estonia, often referred to as the 'Silicon Valley of Europe,' introduced coding in primary schools back in 2012. These moves are not just about creating more software engineers, but about preparing young citizens for a digital-first future. The National Education Policy (NEP) 2020 marked a significant shift in Indian education. One of its key highlights was the introduction of coding from Class 6 onwards. The idea is to nurture computational thinking and expose students to digital tools early. Many CBSE-affiliated schools and private institutions have already started offering basic coding and robotics classes. Online platforms such as WhiteHat Jr, Coding Ninjas, and have capitalized on this trend. A case for making it mandatory We are heading into a world dominated by automation, Artificial Intelligence, and Data Science. Coding is the backbone of all these technologies. Equipping students with this skill early can help India build a future-ready workforce. Coding is not just technical. It's creative. Students learn how to break down big problems into smaller steps—a skill useful in any profession. Even today, the IT industry is one of India's biggest employers. As industries go digital, coding will become essential in finance, healthcare, agriculture, education, and even art. When taught in government schools, coding can empower rural students to access global knowledge, participate in innovation, and break the cycle of poverty. Coding communities are active and supportive. Platforms like GitHub, Stack Overflow, and online courses allow students to learn beyond textbooks. Arguments against Just as not everyone becomes a mathematician despite learning math, not every child will be interested in coding. Forcing it on every student may lead to resistance and disinterest. India has a shortage of trained Computer Science teachers, especially in rural areas. Implementing mandatory coding without proper training can lead to ineffective teaching. Many government schools lack computers, reliable electricity, or internet. Mandating coding without addressing these basic needs will only widen the education gap. With already packed curriculums, adding another compulsory subject could increase the academic burden, especially if it is tested through exams. First, digital literacy, then as core elective Before coding, students must learn basic computer usage, internet safety, and digital ethics. Jumping straight to Python or Java without digital literacy is like teaching literature before learning the alphabet. Instead of making coding mandatory for all, it can be introduced as a core elective. Give students the choice to opt for it early on. Make it mandatory only for those who choose Science or tech-related streams in higher classes. Introduce block-based coding (Scratch) in primary school and text-based coding (Python, C) in high way, we respect individual interests while still preparing a large popu- lation for tech-driven careers. 'Coding should be taught like mathematics—logical thinking first, language second,' says Dr. Ruchi Sharma, a computer science educator in Delhi. 'We need to teach kids to think like coders before we teach them to code,' says Prof. Debarghya Das, a Cornell Computer Science graduate and Google engineer. Impact of coding education Tanmay Bakshi, an Indian-origin coder in Canada, started coding at age 5 and is now an AI expert at age 18. In India, Atal Tinkering Labs funded by NITI Aayog are helping school students build apps and hardware products through hands-on coding experience. It's a misconception that coding is only for those interested in computers. Even in fields like music, medicine, fashion, and agriculture, digital solutions are transforming workflows. Think about doctors using machine learning for diagnosis. Musicians using code to generate beats. Farmers using sensors and coded apps to monitor soil health. Learning to code is like learning a tool that can amplify whatever passion a student already has. Call for smart implementation The question isn't whether coding is important—it clearly is. The real question is how and when we should teach it. Making coding mandatory too early, without preparation, risks alienating students and failing in execution. But delaying or ignoring it will leave the next generation unprepared for a digital economy. The answer lies in balanced reform. Invest in teacher training and infrastructure. Start with basic digital literacy. Gradually integrate problem-solving and logical thinking. Introduce coding as a flexible, project-based subject, not just textbook learning As India's digital economy grows, we must prepare our students to not just use technology—but create it. (Mohsin Iqbal is a student of (Hons) Computer Science at Jamia Millia Islamia New Delhi.)
