What happens when the “golden ticket” of a Stanford computer science education no longer secures a job? Indeed, this is the question haunting top college campuses, and it is one the dear protagonist of The End of Jobs knows far too well. For years, attending one of America’s top colleges meant being poised to enter a job market hungry for smart young engineers. But with the arrival of generative AI, all the preceding rules no longer apply, and not even ‘brainy young programmers’ can get a job, Daniel H. Wilson writes in his rendition of The End of Jobs, which is cited just above.
This generation of senior citizens is familiar with technology and coding that never existed in the time that ChatGPT was initially introduced. To date in 2022, it only has the capability to code for just 30 seconds. Presently, coding done by these robots runs for several hours. Efficient programs are readily developed faster with less bugs compared to that of a junior programmer. “We don’t need the junior developers anymore,” asserted Amr Awadallah, the CEO of Vectara.“The AI now can code better than the average junior developer that comes out of the best schools out there.” he continued.
According to one Stanford study, the hire rate has fallen by about 19% since the end of 2022 for young developers aged between 22-25 yrs. in fields like software development, customer service, and accounting that were exposed to ‘AI’ technology. Whereas fields that have been less exposed to AI technology include nursing, which saw a fall of 13% because students from the University of California, Berkeley; the University of Southern California; and Loyola Marymount University faced the ‘same problem’ in acquiring jobs.
This is the reason why University alum Eylul Akgul was ‘ghosted’ by several hundred companies before landing her first technical leadership role, a role in which she uses ‘AI technology’ but thinks she is working harder than three people combined. The truth about the economics, as leaders in the industry will tell us, says Dario Amodei, the chief executive of Anthropic, is that ‘70% to 90% of the code for some products at my firm will be written by my AI, Claude, within five years, and I expect as much as 50% of all entry-level, white-collar employment to disappear.’ It has happened, says Amy Brooks, vice president of human capital, because “two skilled engineers and one of these LLM-based agents.”
These shifts in organizational structure, says SignalFire, equate to 50% reduction in the number of new role starts for people with under one year of experience within the organization. But in the midst of the shrinking, there are doors opening up for those able to make the transition. AI systems are unable to handle elementary logic, and they must be managed and supervised by humans. Researchers found that experienced programmers will be penalized up to 19% by the AI computer software, since they have to toil harder to validate and correct ‘code produced by machines. The quick functionalization of this supervisory skill-software quality control, attunement to the subtly corrected code, and managing the output of the machines. In the words of Prof. John David N. Dionisio, Experts in ‘computer science’ of our LMU ‘have to supervise and monitor the output of AI, not only execute the AI.
The universities are catching up slowly. Enrollment in fifth year master’s programs have “skyrocketed” at Stanford University as students seek to equip themselves even better in AI before returning to the workforce. But along comes the GenAI in CS Education Consortium for the University of California, San Diego. The group has already begun working on developing an education toolkit in which generative AI is incorporated into computer science education. Google.org is partnering on this to make sure that the next generation of workers understand the world of work that is AI enhanced and that requires knowledge of algorithms, or AI tools, akin to coding.
For young engineers, this means what constitutes “entry level” has to change. The first level of the pole may require highly advanced skills in how AI technologies are governed and how such technology can and should be responsibly implemented. Heather Doshay of SignalFire sums it up well: “The bottom rung is disappearing, but that has the potential to uplevel everyone.”
Just like the generations before them knew how to send and receive emails and surf the internet before they left college and went into the work world, this generation can learn how to effectively implement AI technologies even at the college level. The future will require agility. There are jobs in AI research and development and safety and implementation, and there are jobs in other industries that are not as likely to be automated and include such fields as health care and construction.
Those jobs are on the rise. Others are seeking an education that automatically leads to starting their own companies and developing products that are fashioned from AI and smaller and thus more nimble teams. Others are seeking education that will direct them to be lawyers and counselors and computer science specialist when other industries may not be there. The uncertainty is there, and there is change and opportunity for those who are canny about how to implement and not compete with AI.
