正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

WTH.. did Ox Alpha achieved self-learning, or what ?? I gave it the exact same prompt I used 3 days ago, but the output is completely different. For the same task, it took 330 mins instead of 80 mins and wrote 4,500+ lines of code. how 😬.....?

Adit_Yah ☄️

3,660 subscribers

11,997 次观看 • 4 天前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

WTH, Ox Alpha is literally 10x better today. I gave it the exact same prompt I used yesterday, but the output is completely different. The quality is way better than yesterday. Is it learning continuously or something? 😭

WTH, Ox Alpha is literally 10x better today. I gave it the exact same prompt I used yesterday, but the output is completely different. The quality is way better than yesterday. Is it learning continuously or something? 😭

Bhavy☄️

318,379 次观看 • 5 天前

I feel Ox Alpha has continual learning and is modifying itself each day coz how else does it keep improving on it's own? i gave it the same prompt i used on the first day and the output is drastically improved, with a more detailed and accurate 3d model of raptor engine

I feel Ox Alpha has continual learning and is modifying itself each day coz how else does it keep improving on it's own? i gave it the same prompt i used on the first day and the output is drastically improved, with a more detailed and accurate 3d model of raptor engine

Tim Jayas

542,627 次观看 • 6 天前

🚨 Ox Alpha might actually be AGI I gave it and GPT-5.6 Sol (Ultra) the exact same prompt to build a website with cursor-reactive magnetic motion. Ox Alpha delivered faster and, in this test, far more accurately. the motion feels better, the graphics look cleaner, and even the font choices are way ahead of what GPT-5.6 Sol gave me. I’ll drop the exact prompt I used in the comments for anyone who wants to try it.

🚨 Ox Alpha might actually be AGI I gave it and GPT-5.6 Sol (Ultra) the exact same prompt to build a website with cursor-reactive magnetic motion. Ox Alpha delivered faster and, in this test, far more accurately. the motion feels better, the graphics look cleaner, and even the font choices are way ahead of what GPT-5.6 Sol gave me. I’ll drop the exact prompt I used in the comments for anyone who wants to try it.

m0h

53,425 次观看 • 5 天前

Still, something about it seemed off… So I did a quick google search, and within a couple minutes I found this video of what appeared to be the same woman, with the exact same malady, experiencing the exact same miracle, only in a completely different setting.

Still, something about it seemed off… So I did a quick google search, and within a couple minutes I found this video of what appeared to be the same woman, with the exact same malady, experiencing the exact same miracle, only in a completely different setting.

Matthew Hartke

328,466 次观看 • 3 天前

A sample of caffeine from prompt to launch it took about 30 mins to launch but I wrote an extremely thorough an elaborate prompt. I wanted to QA test if I could crash it. Nope it did QA and solved its own issues for me. My mind was blown.

A sample of caffeine from prompt to launch it took about 30 mins to launch but I wrote an extremely thorough an elaborate prompt. I wanted to QA test if I could crash it. Nope it did QA and solved its own issues for me. My mind was blown.

stoik.icp

21,586 次观看 • 1 年前

I gave ChatGPT a screenshot of a SaaS dashboard and it wrote the code for it. This is the future.

I gave ChatGPT a screenshot of a SaaS dashboard and it wrote the code for it. This is the future.

Mckay Wrigley

8,207,115 次观看 • 2 年前

I recorded this on my phone to practice and so I can listen while i was on the plane ……. This is exactly how I rapped it on the freestyle same exact poccet , but when u go listen to the freestyle it sound completely different

I recorded this on my phone to practice and so I can listen while i was on the plane ……. This is exactly how I rapped it on the freestyle same exact poccet , but when u go listen to the freestyle it sound completely different

SuccaFreeDaloc

28,312 次观看 • 1 年前

🚨 GPT 5.6 Pro first output on the same prompt we are getting started > frontend/ webdev is not solved or improved yet > but understanding increased a lot > it started to take 20-40 mins again like it used to do before 5.5 pro

🚨 GPT 5.6 Pro first output on the same prompt we are getting started > frontend/ webdev is not solved or improved yet > but understanding increased a lot > it started to take 20-40 mins again like it used to do before 5.5 pro

Chetaslua

450,779 次观看 • 2 个月前

It is incredibly sus, how many parts are eerily similar to my own. But, enough of it is completely different to where I can tell you did make it yourself. Although, it is very hard to believe you yourself happened to come up with some of the exact same ideas as me.

It is incredibly sus, how many parts are eerily similar to my own. But, enough of it is completely different to where I can tell you did make it yourself. Although, it is very hard to believe you yourself happened to come up with some of the exact same ideas as me.

Matthew

536,799 次观看 • 6 个月前

Your AI sucks. And it's not the model. It's you. "Make it better" isn't a prompt, it's a guessing game. I changed one step in my process and the output quality completely flipped. Same ideas, same models, totally different results.

Your AI sucks. And it's not the model. It's you. "Make it better" isn't a prompt, it's a guessing game. I changed one step in my process and the output quality completely flipped. Same ideas, same models, totally different results.

Flick

322,001 次观看 • 1 个月前

Figma canvas generation vs MagicPath same prompt. By the time I sent one request to Figma, I had two views and 3 variations of my designs in MagicPath, fully interactive. After 7 mins of waiting for something, I gave up. This is an embarrassing release. Video is 2x speed

Figma canvas generation vs MagicPath same prompt. By the time I sent one request to Figma, I had two views and 3 variations of my designs in MagicPath, fully interactive. After 7 mins of waiting for something, I gave up. This is an embarrassing release. Video is 2x speed

Pietro Schirano

37,689 次观看 • 5 个月前

Trump: That's where Martin Luther King gave his great speech, and he had a million people and I had the same exact crowd, maybe a little bit more. I have pictures of Martin Luther King’s crowd, the exact same, everything, but it was 70 years difference. The exact same crowd, but I actually had more people

Trump: That's where Martin Luther King gave his great speech, and he had a million people and I had the same exact crowd, maybe a little bit more. I have pictures of Martin Luther King’s crowd, the exact same, everything, but it was 70 years difference. The exact same crowd, but I actually had more people

Acyn

464,576 次观看 • 4 个月前

🚨 New Mystery Model - korrine this model is loves a lot of token output even for smaller task , like same prompt for ballister it made the visualization but you can see it made it a website with explanation without asking

🚨 New Mystery Model - korrine this model is loves a lot of token output even for smaller task , like same prompt for ballister it made the visualization but you can see it made it a website with explanation without asking

Chetaslua

36,196 次观看 • 8 天前

🚨 Same sales data. Same prompt. ~32% estimated lower cost. I put the exact same analysis through: Claude Opus 4.8 → $0.12 vs. GMI Router → Cost Mode → Gemini 3.7 Flash → ~$0.081 The task wasn't simple: 6 months of sales data across 4 products → trends, product performance, returns, anomalies, priorities + recommendations. GMI estimated ~$0.0387 savings (~32.3%) based on estimated token usage. And here's what caught my attention: I didn't choose Gemini. GMI Cloud 's Router did. For this workload, the routed output still covered the key analysis requirements. Same data. Same task. Different model selection. Lower estimated cost.

🚨 Same sales data. Same prompt. ~32% estimated lower cost. I put the exact same analysis through: Claude Opus 4.8 → $0.12 vs. GMI Router → Cost Mode → Gemini 3.7 Flash → ~$0.081 The task wasn't simple: 6 months of sales data across 4 products → trends, product performance, returns, anomalies, priorities + recommendations. GMI estimated ~$0.0387 savings (~32.3%) based on estimated token usage. And here's what caught my attention: I didn't choose Gemini. GMI Cloud 's Router did. For this workload, the routed output still covered the key analysis requirements. Same data. Same task. Different model selection. Lower estimated cost.

Altiam Kabir

32,438 次观看 • 2 天前

Opus 4.8 vs GLM 5.2 everyone is praising Glm on my timeline, so i had to test it myself > both models were given the same prompt > GLM was super fast and used like 2% of my weekly limit > Opus took over 20 mins and used like 3% of weekly limits which one do you think won?

Opus 4.8 vs GLM 5.2 everyone is praising Glm on my timeline, so i had to test it myself > both models were given the same prompt > GLM was super fast and used like 2% of my weekly limit > Opus took over 20 mins and used like 3% of weekly limits which one do you think won?

J A Z I I

48,124 次观看 • 2 个月前

I ported Open Brush / Tilt Brush to the open web with IWSDK. It took my agents 29 hours and 3 minutes — I wrote 0 lines of code. GPT 5.5 carried the first stretch; Fable 5 handled the harder work at the end. I did it to prove a point 🧵

I ported Open Brush / Tilt Brush to the open web with IWSDK. It took my agents 29 hours and 3 minutes — I wrote 0 lines of code. GPT 5.5 carried the first stretch; Fable 5 handled the harder work at the end. I did it to prove a point 🧵

Felix Z

19,078 次观看 • 1 个月前

Ox Alpha is behaving weirdly. I gave both of them 3 scenes to build. When the scene is something it's likely seen before, it generates nicely. Give it an unfamiliar scene, and it just doesn't understand it. Kimi K3 nailed all 3 and understood exactly what I was asking for.

Ox Alpha is behaving weirdly. I gave both of them 3 scenes to build. When the scene is something it's likely seen before, it generates nicely. Give it an unfamiliar scene, and it just doesn't understand it. Kimi K3 nailed all 3 and understood exactly what I was asking for.

Bhavy☄️

22,392 次观看 • 4 天前

I gave the same “build me a fully functional Notion-level SaaS workspace” prompt to Marlow, Base44, and Lovable. Base44 → built it, but the result felt more basic, and page-to-page navigation was slower. Lovable → gave me no output on the trial and asked me to add credits. Marlow → built the strongest SaaS of the three, with working interactions, project management, customization, 10 aesthetic themes, and noticeably faster navigation across the generated app. That speed advantage comes from Marlow running the app on your own VM. And it still did all of this on the trial plan. Same prompt. Very different results. See prompt in the comments ↓

I gave the same “build me a fully functional Notion-level SaaS workspace” prompt to Marlow, Base44, and Lovable. Base44 → built it, but the result felt more basic, and page-to-page navigation was slower. Lovable → gave me no output on the trial and asked me to add credits. Marlow → built the strongest SaaS of the three, with working interactions, project management, customization, 10 aesthetic themes, and noticeably faster navigation across the generated app. That speed advantage comes from Marlow running the app on your own VM. And it still did all of this on the trial plan. Same prompt. Very different results. See prompt in the comments ↓

Jaynit Makwana

64,307 次观看 • 15 天前

6 months ago I launched an app now making $300,000 a year I built it 100% by myself all with AI 0 lines of code written Here is step by step how you can do the exact same thing. No secrets held back By the end of the video you'll have everything you need to build with AI:

6 months ago I launched an app now making $300,000 a year I built it 100% by myself all with AI 0 lines of code written Here is step by step how you can do the exact same thing. No secrets held back By the end of the video you'll have everything you need to build with AI:

Alex Finn

36,887 次观看 • 1 年前