Tech2 hrs ago

Oracle Programmers Trained Their AI Replacements, Then Were Laid Off

Oracle forced engineers to train AI that would replace them, then laid them off as AI wrote the company's code, creating buggy output and workforce shifts.

Alex Mercer/3 min/NG

Senior Tech Correspondent

TweetLinkedIn
Oracle Programmers Trained Their AI Replacements, Then Were Laid Off
Source: NewsOriginal source

Oracle compelled its programmers to train AI models that would later write the company's software, then laid them off as the AI took over coding tasks. The shift left senior engineers fixing buggy AI-generated code while junior staff were pushed out.

Context Oracle’s leadership has long argued that building AI infrastructure will dominate the future economy. Chairman Larry Ellison told developers that Oracle’s code is now produced by artificial intelligence, not humans. To meet soaring workloads, the company asked engineers to document their work so the AI could learn. Employees described the task as a no‑win situation: they trained the very system that would replace them while falling behind on deadlines.

Key Facts Workers said they were forced to train AI that would replace them due to overwhelming workload. Oracle’s AI tools now write the code that once required human programmers. Those same AI tools have produced buggy code, wasting senior engineers’ time as they debug and correct the output.

What It Means The episode shows how AI adoption can reshape labor roles even within the firms that create the technology. While automation may increase output, it can also shift skilled work toward verification and maintenance, potentially displacing the original creators. For Oracle, the move signals a broader bet on AI‑driven development, but the immediate cost includes employee turnover and quality‑control challenges.

What to watch next Investors and industry analysts will monitor whether Oracle’s AI‑generated code improves in reliability and how the company balances automation with its workforce strategy.

TweetLinkedIn

More in this thread

Reader notes

Loading comments...