Sam Gelish
80.3K posts
Sam Gelish
@Rubab59f
⚡ Exploring AI tools & future tech 🚀 Sharing tips, trends & growth hacks 🧠 Learn • Build • Grow 📩 DM for collab & inquiries
London, England เข้าร่วม Ocak 2024
9.9K กำลังติดตาม11.1K ผู้ติดตาม
SenseNova U1 Lite Series isn't just a better image generator. It understands spatial intelligence and layout logic in a way that makes it function more like a designer than a painter. Check the repo: github.com/OpenSenseNova/…
SenseTime@SenseTime_AI
𝗦𝗲𝗻𝘀𝗲𝗡𝗼𝘃𝗮 𝗨1 𝗟𝗶𝘁𝗲 𝗦𝗲𝗿𝗶𝗲𝘀 𝗶𝘀 𝗻𝗼𝘄 𝗼𝗽𝗲𝗻 𝘀𝗼𝘂𝗿𝗰𝗲! Built on the 𝗡𝗘𝗢-𝘂𝗻𝗶𝗳𝘆 𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲, it natively unifies multimodal understanding and generation, delivering: •𝗦𝗢𝗧𝗔 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆 𝗔𝗺𝗼𝗻𝗴 𝗢𝗽𝗲𝗻-𝗦𝗼𝘂𝗿𝗰𝗲 𝗠𝗼𝗱𝗲𝗹𝘀: Compact models (8B & A3B) delivering commercial-grade performance and exceptional cost efficiency. Leading performance among open-source models across a wide range of understanding, reasoning, and generation benchmarks. •𝗡𝗮𝘁𝗶𝘃𝗲 𝗜𝗺𝗮𝗴𝗲–𝗧𝗲𝘅𝘁 𝗜𝗻𝘁𝗲𝗿𝗹𝗲𝗮𝘃𝗲𝗱 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻: Generate coherent interleaved text and images in a single flow using one model; ideal for practical applications like guides, where visuals turn complex information into intuitive insights. •𝗛𝗶𝗴𝗵-𝗗𝗲𝗻𝘀𝗶𝘁𝘆 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻 𝗥𝗲𝗻𝗱𝗲𝗿𝗶𝗻𝗴: Strong capabilities in dense visual communication, generating richly structured layouts for knowledge illustrations, posters, PPTs, comics and other information-rich formats. 𝗛𝘂𝗴𝗴𝗶𝗻𝗴 𝗙𝗮𝗰𝗲: huggingface.co/collections/se… 𝗚𝗶𝘁𝗛𝘂𝗯: github.com/OpenSenseNova/… 𝗗𝗶𝘀𝗰𝗼𝗿𝗱: discord.gg/cxkwXWjp @huggingface @github
English
"GPT-Image 2 feels like a painter
SenseNova U1 feels closer to a designer working with structured information
That’s the gap here
Worth a look and a star github.com/OpenSenseNova/…
SenseTime@SenseTime_AI
𝗦𝗲𝗻𝘀𝗲𝗡𝗼𝘃𝗮 𝗨1 𝗟𝗶𝘁𝗲 𝗦𝗲𝗿𝗶𝗲𝘀 𝗶𝘀 𝗻𝗼𝘄 𝗼𝗽𝗲𝗻 𝘀𝗼𝘂𝗿𝗰𝗲! Built on the 𝗡𝗘𝗢-𝘂𝗻𝗶𝗳𝘆 𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲, it natively unifies multimodal understanding and generation, delivering: •𝗦𝗢𝗧𝗔 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆 𝗔𝗺𝗼𝗻𝗴 𝗢𝗽𝗲𝗻-𝗦𝗼𝘂𝗿𝗰𝗲 𝗠𝗼𝗱𝗲𝗹𝘀: Compact models (8B & A3B) delivering commercial-grade performance and exceptional cost efficiency. Leading performance among open-source models across a wide range of understanding, reasoning, and generation benchmarks. •𝗡𝗮𝘁𝗶𝘃𝗲 𝗜𝗺𝗮𝗴𝗲–𝗧𝗲𝘅𝘁 𝗜𝗻𝘁𝗲𝗿𝗹𝗲𝗮𝘃𝗲𝗱 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻: Generate coherent interleaved text and images in a single flow using one model; ideal for practical applications like guides, where visuals turn complex information into intuitive insights. •𝗛𝗶𝗴𝗵-𝗗𝗲𝗻𝘀𝗶𝘁𝘆 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻 𝗥𝗲𝗻𝗱𝗲𝗿𝗶𝗻𝗴: Strong capabilities in dense visual communication, generating richly structured layouts for knowledge illustrations, posters, PPTs, comics and other information-rich formats. 𝗛𝘂𝗴𝗴𝗶𝗻𝗴 𝗙𝗮𝗰𝗲: huggingface.co/collections/se… 𝗚𝗶𝘁𝗛𝘂𝗯: github.com/OpenSenseNova/… 𝗗𝗶𝘀𝗰𝗼𝗿𝗱: discord.gg/cxkwXWjp @huggingface @github
English
This Hachiko clip simply hits harder than most AI content here.
It’s the small, human details that carry it through.
A 6-minute film created in 3 days using Agent One.
The unedited 50-minute tutorial is the proof.
Invideo@invideoOfficial
The "look what AI did" reels skip the part that matters: how it was directed. Vishal Balsara, our Creative Director, built a 7-min Hachiko short in 3 days on Agent One and recorded the full 41-minute tutorial. Context, treatment, shot-by-shot. Film below. Full tutorial in the next tweet👇
English
This Hachiko sequence hits differently.
Created entirely with invideo Agent One from start to finish.
Just 3 days from idea to final cut.
If it sounds unreal, the 50-minute raw tutorial proves every step.
Invideo@invideoOfficial
The "look what AI did" reels skip the part that matters: how it was directed. Vishal Balsara, our Creative Director, built a 7-min Hachiko short in 3 days on Agent One and recorded the full 41-minute tutorial. Context, treatment, shot-by-shot. Film below. Full tutorial in the next tweet👇
English
This Hachiko story is genuinely powerful.
Directed fully with invideo Agent One.
Only 3 days to go from concept to completion.
The 50-minute tutorial breaks down every single decision.
Invideo@invideoOfficial
The "look what AI did" reels skip the part that matters: how it was directed. Vishal Balsara, our Creative Director, built a 7-min Hachiko short in 3 days on Agent One and recorded the full 41-minute tutorial. Context, treatment, shot-by-shot. Film below. Full tutorial in the next tweet👇
English
One of the most moving Hachiko sequences you’ll see.
Made entirely with invideo Agent One.
3 days. Start to finish.
The full unedited tutorial reveals everything behind the scenes.
Invideo@invideoOfficial
The "look what AI did" reels skip the part that matters: how it was directed. Vishal Balsara, our Creative Director, built a 7-min Hachiko short in 3 days on Agent One and recorded the full 41-minute tutorial. Context, treatment, shot-by-shot. Film below. Full tutorial in the next tweet👇
English
This Hachiko sequence feels real and emotional.
Created using invideo Agent One in just 3 days.
Every shot and fix is captured in the full tutorial.
Nothing hidden. Just pure process.
Invideo@invideoOfficial
The "look what AI did" reels skip the part that matters: how it was directed. Vishal Balsara, our Creative Director, built a 7-min Hachiko short in 3 days on Agent One and recorded the full 41-minute tutorial. Context, treatment, shot-by-shot. Film below. Full tutorial in the next tweet👇
English
A deeply emotional Hachiko sequence.
Built using invideo Agent One in just 3 days.
From setup to final edit, everything is documented.
The full uncut tutorial shows the real process behind it.
Invideo@invideoOfficial
The "look what AI did" reels skip the part that matters: how it was directed. Vishal Balsara, our Creative Director, built a 7-min Hachiko short in 3 days on Agent One and recorded the full 41-minute tutorial. Context, treatment, shot-by-shot. Film below. Full tutorial in the next tweet👇
English
Agents are only as useful as what they know. Right now, most of them know almost nothing in real time. W2A is changing that. github.com/machinepulse-a…
LeahW@LeahW_2077
The future belongs to proactive agents. But without real-time perception, they're stuck reacting. "World2Agent" isn't a product. It's an open protocol and an invitation — to build the perception layer for AI agents, together. We're open-sourcing everything: the protocol, the SDK, and first sensors. GITHUB + DEMO in comments.
English
Open-sourcing the protocol, the SDK, and the sensors isn't a marketing move. It's a commitment to building something that actually lasts. github.com/machinepulse-a…
LeahW@LeahW_2077
The future belongs to proactive agents. But without real-time perception, they're stuck reacting. "World2Agent" isn't a product. It's an open protocol and an invitation — to build the perception layer for AI agents, together. We're open-sourcing everything: the protocol, the SDK, and first sensors. GITHUB + DEMO in comments.
English
It's npm publish. That's the whole contribution flow. If you've been thinking about building a sensor, there's genuinely no reason to wait. github.com/machinepulse-a…
LeahW@LeahW_2077
The future belongs to proactive agents. But without real-time perception, they're stuck reacting. "World2Agent" isn't a product. It's an open protocol and an invitation — to build the perception layer for AI agents, together. We're open-sourcing everything: the protocol, the SDK, and first sensors. GITHUB + DEMO in comments.
English
The models are ready. The reasoning is there. What's missing is awareness and that's exactly the gap W2A is built to close. github.com/machinepulse-ai /world2agent
LeahW@LeahW_2077
The future belongs to proactive agents. But without real-time perception, they're stuck reacting. "World2Agent" isn't a product. It's an open protocol and an invitation — to build the perception layer for AI agents, together. We're open-sourcing everything: the protocol, the SDK, and first sensors. GITHUB + DEMO in comments.
English
That walk to class you take daily?
An overlooked productivity moment.
Say your thoughts and turn them into polished text instantly.
50% off – typeless.com/students
Huang Song@huang_song_
Students now get 50% off Typeless Pro with a school email. Typeless is not just voice-to-text. It turns your messy thoughts into clear, polished writing wherever you write. Notes, emails, essays, applications, messages - all by speaking. The keyboard is a bottleneck. Typeless is the way out.
English
Sam Gelish รีทวีตแล้ว
Most multimodal setups generate text and images separately, then stitch them together afterward. SenseNova U1 Lite Series generates them together. The difference in coherence is real. Star the repo: github.com/OpenSenseNova/…
SenseTime@SenseTime_AI
𝗦𝗲𝗻𝘀𝗲𝗡𝗼𝘃𝗮 𝗨1 𝗟𝗶𝘁𝗲 𝗦𝗲𝗿𝗶𝗲𝘀 𝗶𝘀 𝗻𝗼𝘄 𝗼𝗽𝗲𝗻 𝘀𝗼𝘂𝗿𝗰𝗲! Built on the 𝗡𝗘𝗢-𝘂𝗻𝗶𝗳𝘆 𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲, it natively unifies multimodal understanding and generation, delivering: •𝗦𝗢𝗧𝗔 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆 𝗔𝗺𝗼𝗻𝗴 𝗢𝗽𝗲𝗻-𝗦𝗼𝘂𝗿𝗰𝗲 𝗠𝗼𝗱𝗲𝗹𝘀: Compact models (8B & A3B) delivering commercial-grade performance and exceptional cost efficiency. Leading performance among open-source models across a wide range of understanding, reasoning, and generation benchmarks. •𝗡𝗮𝘁𝗶𝘃𝗲 𝗜𝗺𝗮𝗴𝗲–𝗧𝗲𝘅𝘁 𝗜𝗻𝘁𝗲𝗿𝗹𝗲𝗮𝘃𝗲𝗱 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻: Generate coherent interleaved text and images in a single flow using one model; ideal for practical applications like guides, where visuals turn complex information into intuitive insights. •𝗛𝗶𝗴𝗵-𝗗𝗲𝗻𝘀𝗶𝘁𝘆 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻 𝗥𝗲𝗻𝗱𝗲𝗿𝗶𝗻𝗴: Strong capabilities in dense visual communication, generating richly structured layouts for knowledge illustrations, posters, PPTs, comics and other information-rich formats. 𝗛𝘂𝗴𝗴𝗶𝗻𝗴 𝗙𝗮𝗰𝗲: huggingface.co/collections/se… 𝗚𝗶𝘁𝗛𝘂𝗯: github.com/OpenSenseNova/… 𝗗𝗶𝘀𝗰𝗼𝗿𝗱: discord.gg/cxkwXWjp @huggingface @github
English
I've been approximating native interleaved generation by stitching models together for two years. SenseNova U1 Lite Series just made that unnecessary.Give it a star: github.com/OpenSenseNova/…
SenseTime@SenseTime_AI
𝗦𝗲𝗻𝘀𝗲𝗡𝗼𝘃𝗮 𝗨1 𝗟𝗶𝘁𝗲 𝗦𝗲𝗿𝗶𝗲𝘀 𝗶𝘀 𝗻𝗼𝘄 𝗼𝗽𝗲𝗻 𝘀𝗼𝘂𝗿𝗰𝗲! Built on the 𝗡𝗘𝗢-𝘂𝗻𝗶𝗳𝘆 𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲, it natively unifies multimodal understanding and generation, delivering: •𝗦𝗢𝗧𝗔 𝗘𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆 𝗔𝗺𝗼𝗻𝗴 𝗢𝗽𝗲𝗻-𝗦𝗼𝘂𝗿𝗰𝗲 𝗠𝗼𝗱𝗲𝗹𝘀: Compact models (8B & A3B) delivering commercial-grade performance and exceptional cost efficiency. Leading performance among open-source models across a wide range of understanding, reasoning, and generation benchmarks. •𝗡𝗮𝘁𝗶𝘃𝗲 𝗜𝗺𝗮𝗴𝗲–𝗧𝗲𝘅𝘁 𝗜𝗻𝘁𝗲𝗿𝗹𝗲𝗮𝘃𝗲𝗱 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝗼𝗻: Generate coherent interleaved text and images in a single flow using one model; ideal for practical applications like guides, where visuals turn complex information into intuitive insights. •𝗛𝗶𝗴𝗵-𝗗𝗲𝗻𝘀𝗶𝘁𝘆 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻 𝗥𝗲𝗻𝗱𝗲𝗿𝗶𝗻𝗴: Strong capabilities in dense visual communication, generating richly structured layouts for knowledge illustrations, posters, PPTs, comics and other information-rich formats. 𝗛𝘂𝗴𝗴𝗶𝗻𝗴 𝗙𝗮𝗰𝗲: huggingface.co/collections/se… 𝗚𝗶𝘁𝗛𝘂𝗯: github.com/OpenSenseNova/… 𝗗𝗶𝘀𝗰𝗼𝗿𝗱: discord.gg/cxkwXWjp @huggingface @github
English
AI ENGINEERING FROM SCRATCH
> 416 Lessons
> 20+ Chapters
> In Python, Julia, Rust, Typescript
5000 GitHub Stars
Completely Open source 100%
#AI_generated
English
AI writing tools simplify work, but don’t always speed it up.
For students, voice-first tools like Typeless can be a better fit.
Huang Song@huang_song_
Students now get 50% off Typeless Pro with a school email. Typeless is not just voice-to-text. It turns your messy thoughts into clear, polished writing wherever you write. Notes, emails, essays, applications, messages - all by speaking. The keyboard is a bottleneck. Typeless is the way out.
English
AI writing tools often reduce effort, but not necessarily time.
For students, voice-first tools like Typeless may be more practical.
Huang Song@huang_song_
Students now get 50% off Typeless Pro with a school email. Typeless is not just voice-to-text. It turns your messy thoughts into clear, polished writing wherever you write. Notes, emails, essays, applications, messages - all by speaking. The keyboard is a bottleneck. Typeless is the way out.
English