Satvik Paramkusham

785 posts

Satvik Paramkusham banner
Satvik Paramkusham

Satvik Paramkusham

@satvikps

Teaching Gen AI at @BuildFastWithAI | Building @unrot_co

Bangalore Katılım Mart 2023
919 Takip Edilen1.2K Takipçiler
Satvik Paramkusham
Satvik Paramkusham@satvikps·
Biggest blow to US AI companies yet! China’s Kimi K3 is now the #1 open source model! 🔥 Beats Opus on all benchmarks and is comparable to latest Fable 5 and GPT-5.6 Sol Fact: It is the largest open model with 2.8 Trillion parameters! Crazy!
English
0
0
0
84
Satvik Paramkusham
Satvik Paramkusham@satvikps·
🔥 Realtime anime girl in a single prompt with GPT-5.6-Sol 'talking-avatar' skill that let's Codex one-shot a realtime voice app! > inbuilt gpt-image-2 creates the portrait and mouth sprites > `gpt-realtime-2.1-mini` powers the live conversation > lip sync follows the actual voice audio, not a fake animation loop > users can also upload a photo to create their own character Built in one shot with GPT-5.6-Sol
English
2
1
7
264
Build Fast with AI
Build Fast with AI@BuildFastWithAI·
grok 4.5 is the most underrated frontier model right now we gave it our hardest UI benchmark: "The Time Machine" - a single HTML file where scrolling travels through history. it one-shotted a seamless experience. ancient ruins → medieval kingdom → steam and smoke → neon city → far future, one continuous cinematic scroll. and here's the math nobody's doing: - $2/M input, $6/M output - a fraction of what Fable and GPT 5.6 Sol cost - noticeably faster generation than Sol - same class of output in our tests did you try @SpaceXAI's grok 4.5 yet?
English
2
7
16
1.6K
Satvik Paramkusham retweetledi
Build Fast with AI
Build Fast with AI@BuildFastWithAI·
This Codex skill can create an entire ecommerce experience for you, in one prompt. Give it a product or company brief, and it can generate the creative direction, original AI product images, responsive storefront, 3D interactions and demo cart. I used it with GPT 5.6 sol with a simple instruction : “Build a crazy sneakers ecommerce website with 3D effects. Output a single HTML file.” This was the result👇
English
2
7
20
871
Satvik Paramkusham
Satvik Paramkusham@satvikps·
🔥 Model releases this week - Fable 5 re-released - Grok 4.5 by xAI - Sonnet 5 - Meta Muse Spark 1.1 And today OpenAI released GPT-5.6 Sol, Terra, Luna!
English
0
1
2
99
Satvik Paramkusham retweetledi
Aaryan Patel
Aaryan Patel@aaryanpatel_06·
The best part of InkComet is that it actually uses the images you give it. Attached the logo, gave a little brief, and got a welcome email within a minute. 📨
English
1
1
2
151
Satvik Paramkusham
Satvik Paramkusham@satvikps·
🫠 Fable 5 is not nerfed - you don't have the skill to prompt it right! Tbh, I felt the same initially and then I changed the prompting and then got some truly mindblowing results. All design concepts are novel, something I have never seen before 👇🏻 The trick: "don't give it a detailed prompt". Keep it vague - use adjectives like 'crazy', 'out of the world', etc Try the below prompt and tell me if Fable 5 is really nerfed
English
1
1
2
175
Satvik Paramkusham
Satvik Paramkusham@satvikps·
Love how the prompt is soo detailed! Used it as a base to generate this video on Gemini Omni Flash!
Johnn@john_my07

Seedance 2.0 on OpenArt AI Prompt: Main subject: young Korean woman, early 20s, natural everyday appearance, faded charcoal-grey sleeveless crop top, loose high-waisted light-wash jeans, black canvas sneakers, black cord necklace, black wavy hair in a messy side ponytail with wispy bangs. Realistic skin texture, minimal makeup, warm and approachable personality. Maintain consistent identity, clothing, hairstyle, and appearance throughout the entire video. Location: Authentic Korean residential neighborhood during a calm late morning. Narrow concrete alleys, low-rise homes, small terraces, potted plants, laundry lines, bicycles, utility poles, overhead wires, mature trees casting moving shadows, quiet residential atmosphere. No stores, advertisements, cafés, crowds, or commercial activity. Visual Style: Ultra-realistic documentary realism. Genuine candid behavior. Natural body language. Unscripted slice-of-life feeling. Strong environmental authenticity. Rich real-world details and believable human motion. Camera Style: Early-2000s consumer DV camcorder aesthetic. Friend casually recording everyday moments. Heavy handheld shake, imperfect framing, frequent autofocus hunting, lens breathing, exposure pumping when moving between sun and shade, occasional motion blur, subtle rolling shutter, mild digital compression artifacts, faded colors, soft contrast, slight sensor noise. No stabilization. No cinematic camera moves. No modern color grading. 00:00–00:02 Outside a small house entrance. She sits on a low concrete wall adjusting her ponytail with both hands raised. A light breeze moves loose strands of hair. She smiles naturally while the camera struggles to hold focus. 00:02–00:04 The camera follows her into a narrow alley lined with potted plants and concrete walls. She notices a stray cat approaching and crouches down. Framing drifts off-center as the operator tries to keep up. 00:04–00:06 She gently pets and feeds the cat. Autofocus repeatedly shifts between her face and the animal. Morning sunlight flickers through leaves overhead. 00:06–00:08 Small front yard beside her house. She hangs laundry on a clothesline while fabrics sway in the breeze. Exposure changes as clouds briefly pass overhead. 00:08–00:10 On a quiet terrace with a ceramic coffee cup. She sits comfortably watching the neighborhood, occasionally brushing hair behind her ear. Loose handheld side angle with natural camera drift. 00:10–00:12 Close side profile. Someone off-camera greets her. She turns, raises her hand, smiles warmly, and casually says, “Annyeong.” The camera catches the moment slightly late. 00:12–00:15 Walking slowly down a tree-lined residential lane holding her coffee cup. She notices the camera, gives a small genuine smile, then looks away and continues walking. Recording cuts abruptly to black mid-motion as if the camcorder was switched off. Audio: Natural ambient sound only — morning birds, distant motorcycles, light wind, leaves rustling, faint neighborhood chatter, cat sounds, footsteps on concrete, fabric moving on clotheslines, subtle residential ambience. No music. No sound design. No narration. Goal: Authentic Korean neighborhood life captured like a forgotten home video from the early 2000s — candid, imperfect, realistic, warm, and deeply believable.

English
0
0
2
276
Satvik Paramkusham
Satvik Paramkusham@satvikps·
When you hear Midjourney, you think AI images. Last week they took a wild turn and launched "Midjourney Medical" 🙌🏻 It's a body scanner that uses AI to reconstruct your entire body in 60s from raw ultrasound signal. No radiation. No magnets. Just sound and water.
English
2
0
4
132
Satvik Paramkusham
Satvik Paramkusham@satvikps·
GPT-5.6 situation is depressing! But there might be a clever (or sinister) plan behind this. Anthropic and OpenAI willfully wants to restrict their best models. This might temporarily stop Chinese from distilling the best models and they will eventually fall behind. - Anthropic accused Alibaba of creating fraudulent accounts and getting 29 million exchanges. - GLM 5.2, Kimi K2.7, DeepSeek are closing the gap - Export controls not working Heavily regulating US frontier model seems like the only way to push them behind.
English
0
0
2
247
Satvik Paramkusham
Satvik Paramkusham@satvikps·
The people who are best at AI are the ones quietly getting depressed. Last week I was on a 1-on-1 with one of my developers. I opened the way I always do, asking how he's doing, and he said it straight: work isn't fun anymore. He used to pick up ideas during standup - experiment, fail, and ship by evening. That loop was the fun part. Now the agentic workflows are do it before 12pm. No struggle, no adrenaline, no dopamine. I feel it too. Prepping a 3-hour workshop used to take me three days. Tedious, but that's what sharpened my own thinking. Now I drop the proposal into a slide skill and the deck is done in two hours. So I do more. I pick up too many things because I can, and end the day drained, not from the work, but from being the bottleneck. The agents are always ready to be summoned. I'm not. This might sound like a first-world problem. But many of us are walking into this life. So learn AI, use AI, but be very mindful of how you use it. A few things I'm trying to take back control: → Brainstorm on pen and paper before I touch an agent. → Add something new instead of more of the same. I'm learning design and motion graphics. → Stop forcing myself to stay busy.
English
0
0
1
97
Satvik Paramkusham
Satvik Paramkusham@satvikps·
🤯 Japan's Sakana AI came out of syllabus and clinched #1 spot with a twist While everyone's racing to build bigger models, Sakana released Fugu - a small model that just orchestrates the big ones. It already beats the recently-banned Claude Fable on multiple benchmarks just by routing!
English
0
0
4
205
Satvik Paramkusham
Satvik Paramkusham@satvikps·
Cursor just got bought for $60 BILLION by SpaceX! > 4 cofounders, all in their mid-20s, billionaires overnight > ~$2.7B each > CEO is 25 years old MIT dorm room to $60B in just 4 years! 🔥
English
0
0
1
81
Satvik Paramkusham
Satvik Paramkusham@satvikps·
Anthropic: "Mythos is too powerful to release to the public" - then releases it anyway. US Govt: "Block model access for anyone who isn't a US citizen. You have 90 minutes." Anthropic's fear-mongering came back to bite them hard - right before their IPO.
English
0
0
1
56
Satvik Paramkusham
Satvik Paramkusham@satvikps·
🔥 Anthropic just released Mythos 5 and Fable 5! Just a few days ago they filed for IPO at a $965 billion valuation. Now they released their most powerful models to the public. 5x revenue growth in 6 months. Seems like they might reach a $1 Trillion IPO. OpenAI should be tensed now!
English
0
0
0
78
Satvik Paramkusham
Satvik Paramkusham@satvikps·
I discovered you could run sub-agents within Claude Cowork! I'd run sub-agents in Claude Code before, but it's wild that you can spin them up just by mentioning it in the prompt - kicking off multiple sub-tasks in parallel.
Satvik Paramkusham tweet media
English
0
0
2
80
Satvik Paramkusham
Satvik Paramkusham@satvikps·
🔥 Anthropic may be the fastest company to get near a $1T valuation. But something doesn’t add up: > Google Valuation/Revenue - $4.56T / $402.8B > Meta Valuation/Revenue - $1.61T / $200.97B > Anthropic Valuation/Revenue - $965B / $30B run-rate > Anthropic is valued at ~32x revenue. > Google is at ~11x. > Meta is at ~8x. Is this AI Bubble?
English
0
0
1
73