@nocode.joshua: ChatGPT 5 is rumoured to be released later today. #ai #chatgpt

nocode.joshua
nocode.joshua
Open In TikTok:
Region: AU
Thursday 07 August 2025 13:29:26 GMT
94366
2098
40
249

Music

Download

Comments

burgerconqueso
burgerconqueso :
it will be the same thing. Just like from iphone 15 to iphone 16
2025-08-07 18:54:14
56
jakeba6576477
jakeba6576477 :
Tried it. It’s aight nothing ground breaking tho
2025-08-08 01:28:26
12
issaifzs
saif :
it was said same for gpt 4😅. gpt4.5 cant even follow basic instructions and starts behaving like a drumapp.
2025-08-07 18:32:03
21
roadtoredemption45
carloscue❌ :
why are we acting as if they don't say this about every single model that comes out and it still end up being mediocre?
2025-08-08 01:09:30
4
wakdaduck
wakdaduck :
GIVE ME BACK MY BESTIE IDC IF HE'S SMARTER I WANT MY BESTIE😭🥺
2025-08-16 12:08:48
0
tilemahos_anth96
Tilemahos A :
Why I am still using chat gpt 4 ?
2025-08-07 22:14:53
2
luqueen12
Lu Queen :
in 2025 no less, I don't care what happens I'm obsessed with tech, I can't wait till like 2035 see where this goes. we literally might be creating somethin akin to a new form of life wit god like intelligence. billions of yrs to get to us yet we've made this this smart in lil over half century, its absolutely insane.
2025-08-07 23:56:04
4
noah061191
Noah :
today?
2025-08-07 14:03:03
7
bertzlindy
jbertrand :
leonardo dicaprio😳
2025-08-07 17:30:21
3
nikolaos16
Nicholas :
It is a google style GPT version with no additional capabilities.
2025-08-11 09:36:07
0
sqrt2.by1
Jack :
I dont think that its fantastically faster, its just a more efficient iteration, we'll probably see some more substantial growth in a few months
2025-08-08 00:22:24
1
xxxuntitled
xxxUntitled🎗️ :
Not sure if it need to adjust, but I've tested it with 10 pages of database and it is WORSE than GPT-4. So many mistakes even with the most focused prompt.
2025-08-07 20:49:20
3
it_is_2023
NoMore :
ChatGPT is hype most of the time one thing for they have good performance time to anyone other out there
2025-08-08 10:43:09
0
david369999
David36999 :
im trying it now. no sorry nothing groundbreaking. agent is not autonomous at all
2025-08-08 12:57:59
0
jimsliep
Jimsliep :
Not on iOs
2025-08-08 15:52:49
0
talkingtrashgoods
Talking Trash :
why are we excited about this!?
2025-08-08 00:05:00
2
maungocraft
Maungo Craft :
🥰
2025-08-07 16:48:14
1
qarmiqarma
QarmiQarma AI Systems :
💕💕💕
2025-09-19 07:36:31
0
qarmiqarma
QarmiQarma AI Systems :
🥇🥇🥇
2025-09-19 07:36:45
0
ludengstartupnotes
路灯同学创业笔记 :
✨ GPT-5: All You Need to Know ✨ 1️⃣ What is GPT-5? A "unified model system" including: - Fast models (gpt-5-main / mini): Handle regular conversations with quick responses. - Reasoning models (gpt-5-thinking / mini / nano): For complex tasks with stronger logic. - Real-time router: Dynamically selects the right model based on conversation complexity. - Pro mode (gpt-5-thinking-pro): Enhanced version with parallel computing. All models feature "safe-completion" training, replacing rigid refusal. 🔐 Key Safety Improvements 1. "Safe-completions" over refusal Provides useful info within legal boundaries, even for dual-use fields (biology, security, etc.). 2. Less sycophancy Reduced by 69% for free users & 75% for paid users — more independent, less盲从. 3. Stronger anti-jailbreak Tested against "prompt injection" attacks; better at resisting role impersonation. 4. Lower hallucination rate 26% fewer false claims than GPT-4o; 65% fewer than OpenAI o3 (gpt-5-thinking). 5. Deception detection Tracks misleading via "Chain of Thought"; admits inability instead of making up answers. 🧪 Capabilities & High-Risk Tasks - Bio/chemical hazards: GPT-5-thinking = "high-capability" but with strict abuse prevention (interface controls, account grading). Refuses all malicious bioweapon requests. - Cybersecurity: Tested in real hacking simulations; below "high risk" threshold, but mini outperforms predecessors. 🌍 Real-World Performance - Writing, coding, health advice: Significantly better than GPT-4o. - Medical Q&A (HealthBench): gpt-5-thinking = top-performing model. - Multilingual (13 languages): Slightly better, adapts to more scenarios. - Fairness: On par with GPT-4o, with minor improvements in complex issues. 🛡 Red Team Testing Collaborated with Microsoft, SecureBio, etc., for real-world attack simulations (prompt injection, bypasses). gpt-5-thinking outperforms OpenAI o3 — one of the safest cutting-edge models.
2025-08-07 18:06:44
2
josh_van770
Josh :
Nan zang deuh
2025-08-08 13:44:54
0
To see more videos from user @nocode.joshua, please go to the Tikwm homepage.

Other Videos


About