Mastodawn

🚨 A user on X just spotted the first ad popping up inside a live ChatGPT session. The AI‑generated suggestion steered them toward a Peloton deal during a shopping‑research query. Is this the start of in‑chat advertising, or a glimpse of a future GPT‑5 mini feature? OpenAI’s response could shape how we interact with LLMs. Read more to find out. #ChatGPT #OpenAI #GPT5mini #AIgenerated

🔗 https://aidailypost.com/news/user-x-reports-first-ad-appearing-live-chatgpt-conversation

AI Sparkup Nov 26

ChatGPT가 이제 쇼핑 가이드를 만들어줍니다: GPT-5 mini 기반 쇼핑 리서치 출시

OpenAI가 GPT-5 mini 기반의 쇼핑 리서치 기능을 출시했습니다. AI가 대화하며 제품을 조사하고 맞춤형 구매 가이드를 만들어주는 새로운 쇼핑 경험을 소개합니다.

https://aisparkup.com/posts/6980

eicker.news ᳇ tech news Nov 25

#OpenAI launched a new #shoppingresearch feature in #ChatGPT, designed to help users find products and deals. The tool, powered by a version of #GPT5mini, acts as an “expert personal shopper,” creating #personalisedshopping experiences based on user criteria. It offers a #timesaving option for users, especially those on a #budget. https://www.zdnet.com/article/chatgpts-new-shopping-research-tool-is-fast-fun-and-free-but-can-it-out-shop-me/?eicker.news #tech #media #news

ChatGPT's new shopping research tool is fast, fun, and free - but can it out-shop me?

Just in time for the holiday shopping season, OpenAI has introduced a new tool that it hopes will help you find just what you're looking for.

ZDNET

Mmesek Nov 20

Once again, an LLM has a "meltdown". This time GPT-5 Mini started questioning it's own knowledge upon providing a snippet with non existing syntax.

#LLM #gpt5mini #DuckDuckGo #meltdown #AI #GPT

N-gated Hacker News Sep 17, 2025

🤡 Wow, who knew that playing Mad Libs with #AI prompts could magically boost GPT-5-Mini's performance by 22%? 🎩✨ Apparently, a few word swaps are the secret sauce to transforming underwhelming chatbots into overachievers—someone call the Nobel committee. 🚀💥
https://quesma.com/blog/tau2-benchmark-improving-results-smaller-models/ #MadLibs #ChatbotPerformance #GPT5Mini #Innovation #HackerNews #ngated

Tau² Benchmark: How a Prompt Rewrite Boosted GPT-5-mini by 22% - Quesma Blog

We expected small models to be fast, but our benchmarks revealed a common reliability trap. Here’s our deep dive on finding and fixing it.

Quesma Blog

Hacker News Sep 17, 2025

Tau² Benchmark: How a Prompt Rewrite Boosted GPT-5-Mini by 22%

https://quesma.com/blog/tau2-benchmark-improving-results-smaller-models/

#HackerNews #Tau2Benchmark #GPT5Mini #AIResearch #ModelImprovement #PromptEngineering

Tau² Benchmark: How a Prompt Rewrite Boosted GPT-5-mini by 22% - Quesma Blog

We expected small models to be fast, but our benchmarks revealed a common reliability trap. Here’s our deep dive on finding and fixing it.

Quesma Blog

H@R0👨🏻‍💻Aug 22, 2025

用完 #DeepSeekV31 後再回去用 Github Copilot 的 #Gpt5mini 覺得它是智障。 Gpt-5 的每月配額雖然不算少，但從原本的任用變成有配額限制，變相是加價了。

R (rolandinsh)Aug 20, 2025

Pateicoties #GPT5mini īsai pamācībai, beidzot normāli iemācījos taisīt #3d blenderī priekš #DzīvoklisKuldīgā
Šis nepilnos divos vakaros no 0 zināšanām.
Visas iepriekšējās pamācības bija pārāk ... samudžinātas, kur piemirsts, ka kaut kas jāiespējo vai jāaktivizē.
Vēl palika 2 logi un tad jau visas krāsnis.

#blender3d #Blender45