BTMΞDLR retweeted
SMERFs enable real-time, high-fidelity 6DOF navigation in large scenes on standard devices, surpassing current more expensive methods in speed & quality: smerf-3d.github.io/ <--- there's likely an entire side hustle waiting here for those more business/service minded.
Dope. Nice shout out to @LumaLabsAI !
Heads up - new project out from stealth mode.
1
35
BTMΞDLR retweeted
GPTs are a new way for anyone to create a tailored version of ChatGPT to be more helpful in their daily life, at specific tasks, at work, or at home — and then share that creation with others. No code required. openai.com/blog/introducing-…
BTMΞDLR retweeted
Merging streetwear editions with digital.
Flash sale today & tomorrow only: Ordering our Cloudbreaker embroidered packable jacket ($159 USD) unlocks master files to Sky Lounge for customizable, interoperable use across platforms before official release: propagation.house/product-pa…
BTMΞDLR retweeted
Link for PHSDS merch went live last night: propagation.house/category/a… 📡🌐
Additional links & hidden merch can be found at our HQ via Propagation.House or our new games hub, Gridshot.Pro.🫡
Fam, it's time to pour out into the street:
BTMΞDLR retweeted
🫠🤸🏾♂️ Shoutout @gravitysketch, @unity & @Spatial
for offering the tools to take "what if" to "this works" within 24 hours at zero cost. You guys are the future.
#GRIDSHOT started as a morning design sprint & a few weeks later is ending as a finished, cross-platform gameroom:
If you haven't been following the LK-99 "conflama", now might be a great time to tune in - if this all washes out clean, it's another sure fire sign of the singularity on the horizon.
First claimed successful replication of LK-99
Accomplished by a team at the Huazhong University of Science and Technology and posted 30 minutes ago.
Why this is evidence:
The LK-99 flake slightly levitates for both orientations of the magnetic field, meaning it is not simply a magnetized piece of iron or similar 'magnetic material'. A simple magnetic flake would be attracted to one polarity of the strong magnet, and repelled by the other. A diamagnet would be repelled under either orientation, since it resists and expels all fields regardless of the polarity.
Caveats
There is no way to verify the orientation of the strong magnet in this video, also, there are yet to be published experimental measured values of this sample. Diamagnetism is a property of superconductors but without measured and verified data, this is just suggestive of a result.
Take-away
If this synthesis was indeed successful, then this material is easy enough to be made by labs other than the original research team. I would watch carefully for results out of Argonne National Lab, who are reported to be working on their own synthesis of a sample.
This overall corroborates two independent simulation studies that investigated the original Korean authors claim about material and crystal structure, and both studies supported the claims.
Lawrence Berkeley National Lab: arxiv.org/pdf/2307.16892.pdf
Shenyang National Lab: arxiv.org/pdf/2307.16040.pdf
The attached video shows a small flake of their sample responding to an external magnetic field. I scroll through the video to skip to the relevant part.
original video credit to: @altryne
1
649
BTMΞDLR retweeted
Introducing Objaverse-XL, an open dataset of over 10 million 3D objects!
With it, we train Zero123-XL, a foundation model for 3D, observing incredible 3D generalization abilities: 🧵👇
📝 Paper: objaverse.allenai.org/objave…
BTMΞDLR retweeted
Outdoor XR design time-lapse: From freehand architectural draft to real-time global AR location on @Spatial_io 📡🌐🌏
Replying to @bl0ckstone
Digital art doesn't require an indoor setting anymore. Virtual design can happen anywhere. The resulting works can become global, interactive, open access metaverse locations and installations for all. Think about sister cities, street art, hidden communities.
This is huge:
Despite all of the AI talk, keep in mind, we're in the Flintstones stage and the future, will get much weirder still.
Replying to @sharyph_
"Suddenly AI has turned every WiFi router into a camera that can work in the dark, specially tuned to track the living being"
Join for more exciting events like this:
airoad.io/subscribe
41
BTMΞDLR retweeted
Second chapter of #Safehouse coming soon.
This is only a work in progress preview, already working directly in @Spatial_io:
opensea.io/assets/matic/0x29…
Ready for the foundational model wars? @DeepMind's Gemini is gonna be a banger!
Google DeepMind has access to the most complete models of the digital and physical world.
Digital World:
- Search
- Docs
- Drive
- Gmail
...
Physical World:
- YouTube
- Maps
- Waymo
- Photos
...
Brace yourself for the foundational models that will be trained from this treasure trove.
Let’s *just* consider YouTube by itself…
A flabbergasting 500 (!) hours of video is uploaded to the site every single minute.
YouTube is by far the most ubiquitous repository of audio, visual and text data on the internet. And it's not just dance videos - it's tutorials, video essays, podcasts and more. Nothing even comes close.
Conveniently, Google states that Gemini – their next-generation AI model – is "created from the ground up to be multimodal, highly efficient at tool and API integrations, and built to enable future innovations, like memory and planning."
In simpler terms, Gemini is designed to understand (and generate!) different types of data including text, images, and videos, much like how humans process information.
YouTube is a match made in heaven for training a multimodal model like Gemini, especially given its vast content diversity (multiple regions, languages, visual styles, content formats etc).
It’s no surprise that once OpenAI ran out of text data for their hungry hungry hippo, they allegedly ‘scraped’ YouTube and used Whisper transcription to turn audio/video content into even more training data.
Sure, Google could restrict or heavily paywall developers' access to 'scrape' YouTube like other platforms, such as Reddit, have elected to do.
But Google doesn't just have access to the entire corpus of YouTube videos, they also have access to the treasure trove of analytics associated with each video. This metadata on how users engage with a video, the most popular segments, content ID restriction, etc. could prove to be invaluable.
The potential of learning world models from sensory inputs like video is massive, and YouTube’s corpus gives Google an enormous edge.
Of course, OpenAI won’t sit idly by and probably already has mitigation strategies in place to expand their corpus of training data.
And lest we forget, massive video platforms like TikTok, armed with a billion dollars worth of NVIDIA GPUs, are poised to train their own foundational models.
Buckle up, because the AI revolution is just getting started. The peak isn’t even close.
* Enjoyed this post? Drop a like and RT – it keeps the quality content flowing.
* And of course, follow @bilawalsidhu for more AI analysis.
1
52
Deep Minds // 65 billion x zoom
You never know where great ideas come from or who's building new worlds in their mind.
Playing with @midjourney v5.2's new zoom feature. This model + this ability opens up so many doors. Post in AE, music generated by @mubert.app.
1
78
BTMΞDLR retweeted
Meeting a lot of cool people in my DMs.
Seems like way more quality people left here in web3.
I love it.
Metakey is hiring for,
- Project management: (End to end project delivery)
- BDM: (Talk with brands)
- Marketers: (Paid ads, social)
- Sales/Closers: (Talk to brands, show them our products - web3)
- Research: (Need someone to help me personally keep up with innovation, tech, make graphics, break downs, etc). Going hard on Twitter again.
If you're interested in getting a job in web3 read this.
I've hired 40+ for Metakey and this is what I and most others looked for.
(It's worked wonders).
----
1. An obsession and excited about contributing to the space. Usually can see a big smile or pure joy when speaking about their field of skill and how it applies to web3.
2. Not an 'I know everything' attitude but more one of 'If I don't know I'll learn it'.
3. A hobby/background in art, gaming or culture. This space has a lot of it and to truly understand where web3 is going, it's a big bonus to have a passion in relevant areas.
4. A user of products. Bought an NFT? Got burnt? Made some $? joined a DAO? Part of a discord community? Nothing speaks more than actual experience in the trenches.
5. 2yrs+ web3/NFT experience is good but can be compensated for with passion.
6. The ability to work remote, be punctual, be nice and add to the vibe.
7. The ability to wear multiple hats. Most NFT projects/companies are start ups. You'll be doing a variety of jobs/roles.
8. A specialized trained skill set. Art, Marketing, Development, Sales, Project Management. Combine these with web3/NFT experience and you're a weapon.
Yes we're hiring 😅, feel free to DM.
Wanted to see a side by side of @StabilityAI's new SDXL 0.9 vs @midjourney 5.1 running the same prompt... whoa... 👀... July is going to be interesting!
SDXL (left) ----- Midjourney v5.1 (right)
1
1
105

























