Hackernews posts about GPT-3

GPT-3 is a powerful language model developed by OpenAI that uses machine learning to generate human-like text based on the input it receives.

1. LLM Benchmark: Frontier models now statistically indistinguishable
   7 points by js4ever 10 days ago | 4 comments
2. Building an AI cost-optimizer and AI Slop Prevention tool Looking for feedback."
   1 points by mdzakki 19 days ago | 2 comments
3. Ask HN: Has ChatGPT-5.1 Regressed?
   1 points by pedrozieg 21 days ago | 2 comments
4. Open Source AI Reclaims the Digital Commons (gpt3experiments.substack.com)
   6 points by nutanc 2 days ago | discuss
5. Marx's Theory of the "Commons" Explains the AI Crisis (gpt3experiments.substack.com)
   5 points by nutanc 9 days ago | 3 comments
6. ARR is dead. Long live VRR: Vibe Revenue Run-rate (gpt3experiments.substack.com)
   2 points by nutanc about 16 hours ago | 1 comments
7. AI Has a Communism Problem (gpt3experiments.substack.com)
   1 points by nutanc 12 days ago | 2 comments
8. OpenAI is facing every startup's VC question: What if Google copies you? (gpt3experiments.substack.com)
   1 points by nutanc 27 days ago | discuss
9. Show HN: Cumbersome – iOS/macOS API client for OpenAI, Anthropic, Z.ai (apps.apple.com)
   2 points by bluepeter 1 day ago | discuss
10. Show HN: Digital Ski Bootfitter (wayfinderboots.com)
    2 points by brucebotsford 14 days ago | discuss
11. Show HN: NinjaNote – Turn WhatsApp voice notes into searchable, organized notes (www.ninjanote.app)
    2 points by akihoy 19 days ago | discuss
12. GPT-3.5 crashes when it thinks about useRalativeImagePath too much (iter.ca)
    465 points by goranmoomin almost 2 years ago | 151 comments
13. Three Years from GPT-3 to Gemini 3 (www.oneusefulthing.org)
    352 points by JumpCrisscross about 1 month ago | 319 comments
14. The upcoming GPT-3 moment for RL (www.mechanize.work)
    232 points by jxmorris12 6 months ago | 97 comments
15. GPT-4, without specialized training, beat a GPT-3.5 class model that cost $10M (www.threads.net)
    161 points by CharlesW almost 2 years ago | 127 comments
16. Why Google failed to make GPT-3 (www.latent.space)
    41 points by bookofjoe almost 2 years ago | 40 comments
17. Databricks claims its open source foundational LLM outsmarts GPT-3.5 (www.theregister.com)
    26 points by galaxyLogic over 1 year ago | 4 comments
18. OLMo 2 32B: First fully open model to outperform GPT 3.5 and GPT 4o mini (allenai.org)
    25 points by Philpax 10 months ago | 3 comments
19. Show HN: Trained Tiny Tales GPT(30M model)from scratch and deployed in $15 (kunalmishra.info)
    18 points by kunalmishra78 over 1 year ago | 1 comments
20. DeepSeek V2, near GPT4 performance with 30% GPT3.5 cost ($0.14M input tokens) (www.deepseek.com)
    16 points by zinccat over 1 year ago | 2 comments
21. Ask HN: Any good collection of writing prompts for GPT 3.5/4?
    14 points by notabanker almost 2 years ago | 4 comments
22. PawanOsman/ChatGPT: Access GPT-3.5.turbo for free via an API (github.com)
    13 points by gitinit over 1 year ago | 2 comments
23. I made a graph of GPT-3's and GPT-4's price history (twitter.com)
    13 points by yonom over 1 year ago | 1 comments
24. We Built a Language Model 14,000,000x Smaller Than GPT3 and Formally Verified It (github.com)
    9 points by katosteven 5 months ago | 8 comments
25. Azure GPT-3 got replaced by a creepy Chinese state censored model by ByteDance (www.reddit.com)
    9 points by rlevy about 1 year ago | 3 comments
26. Gemma 2B: Scores better than GPT 3.5 Turbo (twitter.com)
    8 points by FergusArgyll over 1 year ago | discuss
27. Outperform GPT-3 with Karpathy's llm.c using just 1/3 training tokens (twitter.com)
    6 points by tosh over 1 year ago | discuss
28. Gemma-2 2B beats GPT3.5 on Chatbot Arena (huggingface.co)
    5 points by danielhanchen over 1 year ago | 1 comments
29. Why Google *couldn't* make GPT-3 (www.latent.space)
    5 points by swyx almost 2 years ago | discuss
30. Free Copilot offering will be downgraded to GPT 3.5 on "peak times" (www.microsoft.com)
    5 points by vitorgrs almost 2 years ago | discuss
31. Signs of consciousness in AI: Can GPT-3 tell how smart it is? (www.nature.com)
    4 points by rntn about 1 year ago | 6 comments
32. Pleading with OpenAI Developers to not retire GPT-3.5-turbo-0613 on June 13th (community.openai.com)
    4 points by yonom over 1 year ago | 1 comments
33. Three Years from GPT-3 to Gemini 3 (www.oneusefulthing.org)
    4 points by dkobia about 1 month ago | discuss
34. Show HN: Code to run Gemma2-2B on a Mac – a powerful model with GPT 3.5 perf (github.com)
    3 points by notnotrishi over 1 year ago | 2 comments
35. GPT-4o mini is significantly smarter and cheaper than GPT-3.5 Turbo (twitter.com)
    3 points by quxinxin over 1 year ago | 2 comments
36. Show HN: Freegptjs, Node.js Free GPT 3.5 API Library (github.com)
    3 points by Dalufishe over 1 year ago | 2 comments
37. Show HN: Built a (mostly) frictionless calorie tracker app (GPT3.5 wrapper) (www.journable.com)
    3 points by stevehoyek over 1 year ago | 1 comments
38. Jailbreak prompt makes GPT 3.5 do anything (www.youtube.com)
    3 points by laurentiurad over 1 year ago | 1 comments
39. NewsCord 2.0 – uses GPT3.5 to find news coverage differences and summarises news (newscord.org)
    3 points by nima_akr almost 2 years ago | 1 comments
40. The upcoming GPT-3 moment for RL (www.mechanize.work)
    3 points by Tamaybes 6 months ago | discuss
41. Tell HN: We've come a long way since GPT-3 days
    3 points by behnamoh 8 months ago | discuss