notes (@notes@social.mikala.one)
當教授講得沒AI清楚：實體教室的危機，與學習本質的再定義
讀者投書 2025-12-14 12:00:00 CST面對AI在知識傳遞上的高效率，傳統單向授課已然失效。大學課堂應轉型為「知識的健身房」，將重心從「講述」轉向「討論、思辨與應用」，培養學生運用工具、判斷資訊並解決真實問題的能力，這才是實體教育的核心價值。
https://www.thenewslens.com/article/262235
#高等教育 #科技 #大學教育 #教育創新 #思辨能力 #學習模式 #實體課堂 #AI #人工智慧 #ChatGPT
Show Original Post
n_dimension (@n_dimension@infosec.exchange)
Train #ChatGpt how to do your job before you get fired.
ChatGPT adds "Skill framework"
Show Original Post
Powerfromspace1 (@Powerfromspace1@mstdn.social)
@fireship.bsky.social
#OpenAI was dead 🪦 … Then GPT-5.2 dropped!
https://youtu.be/rEvEXQvo-F8?si=3pQgJMph4Wo9vZVF
12/13/2025
Show Original Post
rod2ik (@rod2ik@mastodon.social)
Accessible gratuitement sur mobile et le Web, #Euria ne se contente pas de rivaliser avec #ChatGPT sur le plan des fonctionnalités : elle introduit un modèle vertueux (inédit) où chaque requête contribue à chauffer des foyers à #Genève , tout en garantissant qu’aucune donnée personnelle ne sert à l’entraînement des modèles. Ah oui, c’est #opensource !
A tester
https://goodtech.info/euria-ia-gratuite-suisse-alternative-chatgpt-chauffage/
Show Original Post
r (@r@fed.brid.gy)
Accessible gratuitement sur mobile et le Web, #Euria ne se contente pas de rivaliser avec #ChatGPT sur le plan des fonctionnalités : elle introduit un modèle vertueux (inédit) où chaque requête contribue à chauffer des foyers à #Genève. Et #opensource !
goodtech.info/euria-ia-gra...
Désinstallez ChatGPT tout de s...
Show Original Post
zachari (@zachari@aus.social)
Cyber criminals are utilizing a creative attack vector through the use of #AI responses via search engines.
**How they get you:**
What an attacker will do is build a #website and have it added to #LLM responses so when an unsuspecting user googles or opens a conversation with a large language model the attackers website will display in the AI response.
Typically the instructions are easy to follow. Which means even seasoned internet sleuths are being caught out.
**The instructions could**
1. Prompt you to download and install #malware
2. Replicate a popular company or site to sign into which is also known as #phishing
3. Ask for personally identifiable information (PII) to conduct identity fraud
It's a scary thought, considering how many people use #gemini #chatgpt #claude services everyday.
Show Original Post
sylvie (@sylvie@chitter.xyz)
#disnep made it legal for #ChatGPT to draw 3 circles! at this rate, chatGPT will be legally capable of drawing the #PNG logo by 2330!
https://openai.com/index/disney-sora-agreement/
(for the record i dont care, i just wanted to make this joke)
Show Original Post
dnaod (@dnaod@fosstodon.org)
I like #Claude better than #ChatGPT but their login flow is annoying af.
Show Original Post
nat (@nat@kind.social)
Aaahhh this call DID NOT DISAPPOINT! Full recording should be available now. Here's a little excerpt:
“If you're speaking from the edges, or beyond the edges, of empire, agency looks like something else. It's the same with AI. If you're speaking from the edges of empire, and from a culture of resistance, then if something is coming to erase you, or attack you—one of the only options is to get behind how it works, and redirect it.”
https://www.awakin.org/v2/calls/749/vanessa-andreotti
#ai #chatgpt #artificialintelligence
Show Original Post
Bluedepth (@Bluedepth@mastodon.social)
Wow, that was a tour-de-force in awfulness. I had a recipe built in #ChatGPT, it had godforsaken #YAML frontmatter that #Obsidian just refused. (1/3)
Show Original Post
OppaReacts (@OppaReacts@mastodon.social)
ChatGPT in a real robot shows we’re close to disaster. https://youtu.be/byQmJ9x0RWA?si=NGMI36NXsAYmihnx
#ai #technology #existential #extinction #technology #skynet #terminator #robots #chatgpt #disaster #humanity #death
Show Original Post