The Tech Industry pays programmers handsomely to tap the right keys in the proper order. Nevertheless, earlier this month, entrepreneur Sharif Shameem tested an alternative way of writing code. Initially, he wrote a short description of a simple application for the purpose of adding items to a to-do list and check them off when completed. Afterward, he submitted it to an artificial intelligence system called GPT-3.
GPT-3 digested large swaths of the web, including tutorials. Then, later, the system spat out some functioning code. Shameem says he got chills down his spine. He was like, ‘Woah, something is different.’
The Research lab OpenAI created GPT-3. Moreover, GPT-3 is currently provoking chills across Silicon Valley. Last month, the company launched the service in a beta mode. From then on, it has gradually widened access. The ceremony went viral among investors and entrepreneurs in the past week. They excitedly took to Twitter to discuss and share results from prodding GPT-3 to generate guitar tabs, tweets, poems, and memes.
This viral software event is an experiment in what happens when new artificial intelligence research is placed and packaged in the hands of tech-savvy people. Nevertheless, they are not artificial intelligence experts. The system of OpenAI has been feted and tested in ways it did not expect. The results show the potential use of technology. Nevertheless, it alsi shows its limitation – and how it can lead people astray.
Videos of Shameem show that GPT-3 responds to prompts like ‘a button that looks like a watermelon’ by coding a pink circle with a green border and word watermelon. It went viral and prompted gloomy forecasts concerning the employment prospects of programmers.
Delian Asparouhov is an investor at Founders Fund, an early backer of Facebook and SpaceX cofounded by Peter Thiel. He blogged that GPT-3 provides 10,000 PhDs. Those PhDs are willing to converse with you.
Asparouhov fed GPT-3 the start of a memo on a prospective investment in health care. The system added a discussion on regulatory hurdles and wrote that it would be comfortable with that risk. This is because of the massive costs (sic) and massive upside savings to the system.
Moreover, other experiments have explored more creative terrain. Elliot Turner is a Denver entrepreneur. He said that GPT-3 could rephrase rude comments into pleasant ones. Or it can insert results to be vice versa. Gwern Branwen is an independent researcher. He generated a trove of literary GPT-3 content, including pastiches of Harry Potter in the styles of Jane Austen and Ernest Hemingway.
Directing machine-learning built GPT-3 to study the statistical patterns in almost a trillion words collected from digitized books and the web. Thus, the system memorized the forms of countless situations and genres. This happened in from sports writing to C++ tutorials. It uses its understanding of that immense corpus for responding to a text prompt by generating new text with similar statistical patterns.
The results appear to be technically impressive. They can also be thought-provoking or fun, as the code, poems, and other experiments attest. GPT-3 reliably repeated the format and combined precise details like past employers with fabrications like a deadly climbing accident.