 |
| OpenAI Improves Code-Generating AI |
Table of Contents
OpenAI's AI Codex: Revolutionizing Code Generation with Natural Language Processing
ChatGPT OpenAI AI industry mein aik wabaal ban raha hai, lekin company wahan rukti nahi hai. OpenAI apnay AI Codex ko barqarar rakhnay mein laga rehta hai, jo natural language instructions ko kaam kartay code mein tabdeel karta hai aur uski efficiency mein koi muqabla nahi hai. AI technology ki taraqqi ke sath, OpenAI ka AI Codex developers aur businesses ke liye aik mustahiq hal ban jaega, code generation process ko streamline karke aur AI aur programming ke khandan mein naye imkaanon ko kholtay huay.
OpenAI's ChatGPT: Threatening Google with Advanced Natural Language Chatbot
OpenAI ne haal hi mein apnay natural language chatbot ChatGPT ke sath khabron mein shor machaya hai, jo itna taraqqi kar chuka hai ke ab yeh Google ko khatray mein daal raha hai. Lekin ChatGPT OpenAI mein in kay bohat se products mein se aik hai. Sab se zyada ummidwar AI products mein se aik Codex hai, lekin isay kam attention milti hai kyunki yeh specifically developers ke liye design kiya gaya hai.
Codex's Goal: Generating Code from Natural Language Requests
Codex ka maqsad user ki natural language requests se computer code generate karna hai. Programming mein, Codex ka bohat bara asar ho sakta hai, kyunke yeh companies aur developers ko applications tezi se ya kam programmers ke sath create karne ki ijazat deta hai. Codex pehlay se hi istemaal mein hai, kyunke yeh Microsoft ke Github platform ke Copilot feature ke peeche ka AI hai (OpenAI ke investor).
Unlocking the Power of AI: Understanding the Technology Behind GitHub's Copilot
Github par, Copilot code likhne ke liye autocompletion ke tarah kaam karta hai. OpenAI technology ko behtar kar raha hai aur Semafor article ke mutabiq, ek taamer-e-moazzam se kaam kar raha hai jo ki contracted developers (na ki employees) ke sath kaam karta hai Codex ko optimize karne ke liye. Aaj, OpenAI tech giants ke muqable me kafi chota hai, sirf 375 employees ke sath, lekin AI ko train karne ke liye zaroori tasks ke liye woh contractors par kafi rely karta hai. Haqeeqat me, Semafor riwayat karta hai ki OpenAI ne haal hi me apne products ko majboot karne ke liye 1,000 contractors ko taamer kiya hai.
In naye contractors ki 60% AI ko khurak dene ke liye data labeling kar rahe hai, jabki 40% Codex ko train kar rahe hain. Pehle, Codex GitHub data ke sath introduce kiya gaya tha, lekin ab OpenAI natural language se AI training ke liye specifically code aur comments ka istemal karna chahta hai. "Saral natural language commands ko interpret karne me saksham aur istemal karne ke liye user ke naam par execute karne me saksham, Codex ab dher sari programming languages ko samajh sakta hai, jo existing applications ke liye natural language interface ko enable karta hai", OpenAI ke Codex ke bayan ke mutabiq.
The Role of Contractors in OpenAI's Codex Development: An Analysis
Semafor ke zikr kiye gaye ek developer ke mutabiq, OpenAI ke test ke dauran, unhe sabse pehle ek problem diya jata tha jisko code se solve karna tha, aur unhe bataana tha ki woh is problem ko kaise approach karna chahte hain English me. Phir doosre phase me, unhone code likha. Aur jab is code me bug detect hua, to unhe iske baare me ek text likhne ke liye kaha gaya tha ki code ko kaise correct karna hai. "Unhe shayad is model ko ek khaas tarah ke training data se khurak dena chahte hain, jahan insan apne thought process ke step-by-step presentation ko provide karta hai", is developer ne bhi Semafor ke zikr kiya hai.
Beshak, yeh jaankari official banani chahiye. Lekin yeh bhi dhyan me rakhna laayak hai ki OpenAI model development ke dauran in tasks ko perform karne ke liye contractors par bharosa karta hai. Misal ke taur par, company ne in contractors ko ek scientific paper me dhanyavad diya hai jo 2022 me publish kiya gaya tha. "Akhir me, hum apne sabhi contractors ka shukriya ada karna chahte hain jo models ko train karne ke liye zaroori data provide kiye hai," is publication me likha tha.