nathan lile Profile Banner
nathan lile Profile
nathan lile

@NathanThinks

Followers
1,838
Following
938
Media
216
Statuses
1,430

ceo/cofounder @ (we're hiring!) #GenerativeAI recurrent rabbit hole victim. swims in data lakes & pools. nothing great is easy.

San Francisco
Joined August 2013
Don't wanna be here? Send us removal request.
Explore trending content on Musk Viewer
Pinned Tweet
@NathanThinks
nathan lile
4 months
Excited to announce that I've founded new company! is dedicated to solving the AI alignment challenge–ensuring advanced AI systems are robustly aligned with human intentions & values Honored to have @lcastricato @fdesouza as cofounders & Seed funding from
Tweet media one
5
6
60
@NathanThinks
nathan lile
1 year
Stable @ClipDropapp 😊 "Stability AI’s First Acquisition Furthers Its Goal Of Delivering The Most Advanced Creative Applications With Its Open Source Generative AI Models" Read more ➡️
9
101
448
@NathanThinks
nathan lile
1 year
🚀 New Stability AI @Blender Addon🚀 Use #StableDiffusion inside your favorite 3D software. Generate textures, generate AI video from your renders, and more. ❤️ #b3d
3
51
191
@NathanThinks
nathan lile
1 year
Woahhhh... wtf?! "No Modality Left Behind" = transform ANY user input into high-def generative #GenerativeAI audio with Make-An-Audio 🎧 📝 Text → 🎵 📸 Image → 🎵 🎥 Video → 🎵 🖌️ Inpainting → 🎵
4
33
188
@NathanThinks
nathan lile
1 year
📸 Take a depth scene screenshot of camera view 🎨 Use Controlnet to generate images with the screenshot 💻 Import the texture and project it from the same camera view 🚀 #UE5 + #StableDiffusion2 FTW
1
20
99
@NathanThinks
nathan lile
1 year
AI unlocks creativity by flattening learning curves (scripting 🧑‍💻) and eliminating unnecessarily tedious tasks. Your own personal troubleshooter & liaison—Asking ChatGPT to write scripts for complex 3D tasks in @Blender . Terrain, textures, animations... 3 point lighting!
3
12
89
@NathanThinks
nathan lile
1 year
🍿 De-aging Harrison Ford with Stable Diffusion 🍿 SD deepfakes = "the next autotune of the entertainment industry"
6
22
82
@NathanThinks
nathan lile
1 year
🚀 We just launched SDXL 0.9 with massive improvements in image & composition detail. 3.5B base model & 6.6B model ensemble pipeline + 2x CLIP models, including the largest OpenCLIP model to date. Research blog coming soon.
Tweet media one
Tweet media two
5
6
40
@NathanThinks
nathan lile
7 years
@kylegriffin1 Y'all are all mindlessly jumping on a freak-out bandwagon. "Verified" isn't some preferential endorsed by @Twitter status, it's just proof that an account is representing who they claim to be. You *want* well known ppl to be verified so you're not tricked by fake accounts.
9
4
45
@NathanThinks
nathan lile
1 year
😉
Tweet media one
@EMostaque
Emad
1 year
Tweet media one
58
289
2K
2
11
49
@NathanThinks
nathan lile
1 year
💯 🔥BLIP-2🔥 demo surpassed my wildest expectations @LiJunnan0409 @_akhaliq
Tweet media one
@LiJunnan0409
Li Junnan
1 year
🔥BLIP-2🔥 demo is live! Come play with LLMs that can understand images and share your examples! Project page: BLIP-2 knows mass–energy equivalence! More examples in the 🧵
Tweet media one
26
187
732
2
12
49
@NathanThinks
nathan lile
1 year
Quilling Starry Night with #DeepFloydIF #DeepFloyd
Tweet media one
Tweet media two
Tweet media three
@NathanThinks
nathan lile
1 year
IF I quilled, I'd use #DeepFloydIF to ideate first, before spending weeks to create a very involved design. #AIart #DeepFloyd
Tweet media one
Tweet media two
Tweet media three
Tweet media four
3
3
30
2
3
48
@NathanThinks
nathan lile
2 years
The first artists🎨 who embrace open source #ai are gonna see ridiculous network effects 📈
8
5
45
@NathanThinks
nathan lile
1 year
"Death will come, and she will have your eyes" Took me a while 🙈 @_bra_ket #DeepFloyd #DeepFloydIF
Tweet media one
10
7
43
@NathanThinks
nathan lile
2 years
@Timcast These are hilarious, but not reflective of modern AI text-to-image. If you haven't seen DALLE, I made a list that will probably blow your mind:
1
1
1
@NathanThinks
nathan lile
1 year
Technical illiteracy is at the core of 90% of the generative #ai / #AIart hysteria.
@DiffusionPics
Stable Diffusion 🎨 AI Art
1 year
Let's be clear here - Using AI does not break laws, it is free - it is not a corporate monopoly, it is not fascism, it is not theft, it is no plagiarism, and it certainly is not ok to wish or imply that you hope that us "tech bros" die in a hail of bullets. #AIArt #StableDi
Tweet media one
26
30
216
7
6
41
@NathanThinks
nathan lile
8 years
#Omaha isn't ready for us #TwoStrong
Tweet media one
1
3
39
@NathanThinks
nathan lile
1 year
Scaling generative AI with @awscloud @nin_artificial explains why @StabilityAI partnership with AWS is crucial for new open source foundation models 🚀
3
9
39
@NathanThinks
nathan lile
7 years
I'M GLAD THEY TOLD ME THEY MISS ME, BECAUSE ITS **CLEAR** THEY DIDN'T NEED ME -- This program won't stop improving #FLIPPEDTHESWITCH
Tweet media one
0
6
37
@NathanThinks
nathan lile
1 year
The Spotify model: autonomous "squads" for focused projects, group related squads in "tribes" for collaboration, organize "chapters" by skill sets, and "guilds" to foster a culture of trust, innovation, and continuous learning
Tweet media one
1
1
8
@NathanThinks
nathan lile
2 years
Anticipating an Airbnb competitor that, ironically, targets the the company's early origins—poors willing to sleep on an air mattress.
0
1
3
@NathanThinks
nathan lile
1 year
AI video model (gif printer go brrrr) by the brilliant minds at @StabilityAI @andi_blatt @robrombach @timudk Predicting: the gift that keeps on gif-ing! 🎁
@_akhaliq
AK
1 year
Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Models turn the publicly available, state-of-the-art text-to-image LDM Stable Diffusion into an efficient and expressive text-to-video model with resolution up to 1280 x 2048 abs:
26
349
1K
2
5
32
@NathanThinks
nathan lile
11 months
After an amazing year at @StabilityAI , I'm excited about a new chapter. This past year has been nothing short of amazing, a testament to the extraordinary people who I've had the privilege to work with. I want to extend my heartfelt gratitude to @EMostaque and the entire
3
0
32
@NathanThinks
nathan lile
2 years
Tweet media one
1
2
30
@NathanThinks
nathan lile
1 year
"Confronting Free Will" in #SDXL
Tweet media one
2
2
28
@NathanThinks
nathan lile
1 year
IF I quilled, I'd use #DeepFloydIF to ideate first, before spending weeks to create a very involved design. #AIart #DeepFloyd
Tweet media one
Tweet media two
Tweet media three
Tweet media four
3
3
30
@NathanThinks
nathan lile
9 years
Proof that a 4.0 means nothing: I forgot to bring a towel to the Nationals swim meet this morning #Phillips66Nats
2
3
27
@NathanThinks
nathan lile
1 year
#DeepFloyd "IF" code is live! 🎉 🔜 In the coming days expect: - Model weights 🧠 - Blog posts 📝 - Inference & tools 🔧 - Research & more! 🎯 🥇 Zero-shot FID-30K score of 6.66 (Beats Imagen, Dall-E 2, Parti, eDiffi, Stable Diffusion benchmarks)
@NathanThinks
nathan lile
1 year
Quilling The Dark Side of the Moon 🌚 requested by @danberridge 😊 #DeepFloydIF #DeepFloyd @deepfloydai
Tweet media one
Tweet media two
Tweet media three
Tweet media four
4
2
16
3
5
29
@NathanThinks
nathan lile
1 year
Cherry-picked AI images made with DeepFloyd IF #deepfloyd #deepfloydif
Tweet media one
Tweet media two
Tweet media three
Tweet media four
2
1
28
@NathanThinks
nathan lile
9 years
I can do all things through Advil who numbs me. #blessed
0
3
22
@NathanThinks
nathan lile
10 months
Medusa can improve the LLM generation efficiency by 2x Instead of using an additional draft model like speculative decoding, Medusa merely introduces a few additional decoding heads
2
3
24
@NathanThinks
nathan lile
1 year
🚨 BREAKING: @DeepFloydAI allegedly trained exclusively on really good memes & psychedelic space-rock. #DeepFloyd #DeepFloydIF
Tweet media one
@deepfloydai
DeepFloyd
1 year
If I were a swan, I'd be rainbow-colored. If I were a train, I'd be intergalactic. If I were a model, I'd be open-source. @AVShonenkov , @_bra_ket , @_gugutse_ , @susiaiv , @StabilityAI #deepfloyd #deepfloydif
Tweet media one
Tweet media two
Tweet media three
Tweet media four
35
94
557
1
1
22
@NathanThinks
nathan lile
1 year
Just 1 day since #StableLM launch, & GitHub stars already skyrocketing 🌟 To prevent Y-axis congestion, GPT-NeoX, Pythia etc left out. Featuring LLaMA, Alpaca, Dolly, & Vicuna below. Adoption is accelerating, but this is just the start - Buckle up, folks! 🚀
Tweet media one
2
8
23
@NathanThinks
nathan lile
8 years
😲 @Nike
1
4
20
@NathanThinks
nathan lile
9 months
@AravSrinivas "cringe territory"
Tweet media one
0
0
22
@NathanThinks
nathan lile
1 year
"Open AI"
Tweet media one
4
6
22
@NathanThinks
nathan lile
4 months
congratulations to @DivGarg9 @omarshaya & the @MultiOn_AI team on hitting 500K agent queries 🎉🚀
Tweet media one
1
5
20
@NathanThinks
nathan lile
1 year
"Frosted Doodles for Breakfast" Wayyy too cute to eat 🥰 #DeepFloydIF #DeepFloyd #AIart
Tweet media one
Tweet media two
2
3
19
@NathanThinks
nathan lile
1 year
Late, but couldn't help myself. #DeepFloyd is great at AI caricatures. "This new Bing will make Google come out and dance, and I want people to know that we made them dance" — @satyanadella #DeepFloydIF
Tweet media one
5
2
19
@NathanThinks
nathan lile
1 year
spaghet cat
Tweet media one
2
1
19
@NathanThinks
nathan lile
1 year
joe & team's beta #sdxl is ridiculous y'all
Tweet media one
Tweet media two
2
5
19
@NathanThinks
nathan lile
1 year
Quilling The Dark Side of the Moon 🌚 requested by @danberridge 😊 #DeepFloydIF #DeepFloyd @deepfloydai
Tweet media one
Tweet media two
Tweet media three
Tweet media four
@NathanThinks
nathan lile
1 year
Quilling Starry Night with #DeepFloydIF #DeepFloyd
Tweet media one
Tweet media two
Tweet media three
2
3
48
4
2
16
@NathanThinks
nathan lile
2 years
“No, sir, we did not use copyrighted images. We believe Dall-e 2 spent too much time on the internet and has learned to enjoy trolling.” (Alamy stock image watermark) #dalle2 #dalle #gpt3
@Dalle2Pics
Best Dalle2 Pics
2 years
a transparent sphere on a beach with a crab looking at it #dalle2 #dalle
Tweet media one
7
30
546
1
2
8
@NathanThinks
nathan lile
8 years
A pic of my butt made it into UGA's Women's Team group message. Apparently not many butts make it. I think I just hit my lifetime peak.
0
3
15
@NathanThinks
nathan lile
9 months
I disagree. Apple & OpenAI might too. Apps with no value-add features, just profiting from "AI Chat" keywords, are competition magnets 🧲. Go make a better app (yes, 🫵). As competition grows 📈, they experiment with positioning and marketing, accelerating the potential to
Tweet media one
@Scobleizer
Robert Scoble
9 months
Apple shouldn’t allow this.
69
16
239
0
2
6
@NathanThinks
nathan lile
1 year
Tweet media one
1
2
15
@NathanThinks
nathan lile
2 years
Now that Twitter knows about #dallemini , y'all might as well have fun with these other free generative AI apps. A few 💎's you might not know about...👇
1
3
15
@NathanThinks
nathan lile
1 year
Tweet media one
1
2
15
@NathanThinks
nathan lile
8 years
Just slept through my last undergrad class ever. Yes, the one that started at 11:50.
0
1
14
@NathanThinks
nathan lile
1 year
Playing with Animai, the newest free generative ai tool by @RevelXyz @didisimon 💜💁🏼‍♂️
14
4
15
@NathanThinks
nathan lile
2 years
@mcuban @elonmusk @Meta Twitter → Global consciousness/information. Until we all get @elonmusk 's Neuralink, written text is the most efficient way to consume information. AR/VR on @Meta is unlikely to directly compete with @Twitter 's strength.
0
0
3
@NathanThinks
nathan lile
2 years
#StableDiffusion doesn’t restrict creativity
2
0
13
@NathanThinks
nathan lile
2 years
@levelsio @stripe This is such a cool live case study, thanks for building in public 💯
0
0
12
@NathanThinks
nathan lile
2 years
@bowtiedbower @GuruAnaerobic Maybe a better way to frame this would be, "one piece of specific advice is difficult to generalize or systematize." If you blindly follow a piece of specific advice... good luck iterating.
0
1
11
@NathanThinks
nathan lile
1 year
Goldman predicts generative AI could increase global GDP by 7% and lift productivity growth by 1.5 percentage points over a 10-year period. Estimate that two-thirds of US occupations are exposed to some level of automation by AI, of which between a 1/4 and 1/2 of their workload
Tweet media one
1
4
13
@NathanThinks
nathan lile
1 year
#deepfloyd weights are out! 🤖 Introducing DeepFloyd IF: a modular, cascaded, pixel diffusion model for high-res text-to-image synthesis. 🖼️ Trained on 1B (image, text) pairs, it holds potential for art, design, storytelling, VR & more! Try it! ➡️
@StabilityAI
Stability AI
1 year
🚨Announcing the release of DeepFloyd IF🚨 Our multimodal AI lab, @DeepFloydAI , is publicly releasing their state-of-the-art text-to-image model. Learn more here →
Tweet media one
Tweet media two
Tweet media three
Tweet media four
27
377
1K
0
4
13
@NathanThinks
nathan lile
2 years
@AntonNellCrypto Hey Anton, I’d like to continue maintaining/operating chainlist—would you be willing to hand over ownership instead of letting the domain expire & get snapped up at auction by a phisher?
1
0
11
@NathanThinks
nathan lile
10 years
Our first encounter with the homeless in Baltimore. #ccsa http://t.co/J7MVgDkyyt
Tweet media one
1
5
12
@NathanThinks
nathan lile
2 years
@naval Fair/unfair is a low-utility lens for reality
2
1
13
@NathanThinks
nathan lile
9 years
Tweet media one
0
2
10
@NathanThinks
nathan lile
1 year
Tweet media one
2
0
12
@NathanThinks
nathan lile
9 years
Tweet media one
1
2
10
@NathanThinks
nathan lile
9 years
Congratulations @UMBCSwimDive on a fantastic conference & overall season. I couldn't be happier yall joined our conference
0
1
11
@NathanThinks
nathan lile
1 year
🤖 Demo of StableLM tuned Alpha 7B by @StabilityAI Try the @huggingface space! ⬇️⬇️⬇️
0
1
11
@NathanThinks
nathan lile
1 year
Massive AI models (like GPT-4) aren't an efficient or scalable approach. Smaller & more specialized AI models working together or in a decentralized approach is a more viable solution 🌐🤖
@EMostaque
Emad
1 year
Sure giant models aren’t the way but swarms are but like was google ever going to give giant models to billions of users? Would need all the flops in the world 🌎 Google’s core business is aggregation per @stratechery innovation etc is cherry on top
3
3
69
1
2
11
@NathanThinks
nathan lile
1 year
"The portfolio already includes investments in AI startup darlings OpenAI, Stability AI Ltd. and Anthropic."
Tweet media one
1
2
11
@NathanThinks
nathan lile
7 years
What does it mean if the Holy Water sizzles when it hits your skin? (Asking for a friend)
0
0
10
@NathanThinks
nathan lile
9 years
Yo @GWU_Tech_Serv all my blackboard courses just evaporated #Day1OfFinals #gwupride
0
1
6
@NathanThinks
nathan lile
7 months
In just 3 months, @SakanaAILabs (co-founded by my friend @hardmaru ) lands NTT partnership, the world's 4th largest telecom 🚀 They're building an AI constellation—a distributed net of smol, task-oriented AIs, autonomously cooperating A big step towards collective intelligence!
@SakanaAILabs
Sakana AI
7 months
We’re excited to be partnering with NTT Group to develop AI in Japan! Partnership announcement on @Nikkei NTT、元Google研究者が設立のサカナAIと提携 NTT partners with Sakana AI, a company by former Google researchers.
3
49
306
0
0
10
@NathanThinks
nathan lile
1 year
How many years until 🔞/bias risks of #GenerativeAI are implicit & largely ignored? Like with web browsers.
2
0
5
@NathanThinks
nathan lile
9 years
It's a good thing that changing quickly and pronouncing words aren't necessary qualities in beauty pageants #missgwu2015
0
3
10
@NathanThinks
nathan lile
2 years
I think I'm in love 💚 @pipedream Adding Python code (serverless) between generic (pre-built) App/API tasks + /tmp access + generous dev tier big enough to experiment = immediately solved EVERY pain point I had with Zapier/IFTTT/Et al. (Fantastic brand name too)
Tweet media one
2
4
9
@NathanThinks
nathan lile
9 months
@willdepue I'm hoping Siri gets a MS Paint level "hell freezes over" update
Tweet media one
0
0
8
@NathanThinks
nathan lile
1 year
@clipdropapp ClipDrop just shipped ControlNet + brand new Upscaler + HQ Cleanup!
@sudu_cb
Sudharshan
1 year
ControlNet + @clipdropapp API's have been integrated! You can now use much better normal maps to capture the juiciest features with Stable Diffusion. @EMostaque
Tweet media one
8
67
542
1
1
9
@NathanThinks
nathan lile
8 years
0
2
9
@NathanThinks
nathan lile
2 years
my @StabilityAI canned response ✉️ 📉 “Sry, we've been...🧪💣🚀” 📉 “Oops” 📈 "The #ai accidented & sent the #chatGPT w/o asking me 1st" 📈 “Indelibly confusing email miscommunications like this are actually valuable training data 😮☝️!”
@JamesIvings
James Ivings
2 years
It's ready! 🚀🚀🚀🚀🚀 Introducing Ellie - your email assistant! 💌 Ellie is powered by @OpenAI and will learn your writing style and reply to emails as if you wrote them 🔥 If you want to be an early user, please retweet and comment below and I'll send you an invite code! 🙂
2K
2K
6K
0
1
9
@NathanThinks
nathan lile
1 year
nobody: gen ai: anything🤝everything chatgpt: CoDi-fied masterpiece 🥹
0
1
8
@NathanThinks
nathan lile
9 years
Today I won naming rights to @tatertot075 's 7th offspring.
1
1
7
@NathanThinks
nathan lile
9 years
Tweet media one
0
1
8
@NathanThinks
nathan lile
2 years
@anatoly Just checked TikToks permission request, no clue why this isn’t scoped properly.
Tweet media one
0
2
8
@NathanThinks
nathan lile
1 year
ChatGPT stock portfolio outperforming top UK investment funds 19% of UK adults would consider getting financial advice from ChatGPT A further 8% said they were already doing so... 🫠 Yet, 38% of Brits still ponder, "What's a ChatGPT?" 🤔
Tweet media one
1
0
8
@NathanThinks
nathan lile
1 year
i'm pretty good at gaslighting robots now thank you, chatGPT
1
0
7
@NathanThinks
nathan lile
10 years
Tweet media one
0
1
7
@NathanThinks
nathan lile
9 years
"They didn't know that it was impossible, thus they did it." - Mark Twain
0
1
7
@NathanThinks
nathan lile
10 years
"I think I just found a Canadian. He thanked me for walking through the door he opened." - Luke
0
1
7
@NathanThinks
nathan lile
1 year
Private data > Public data The future = lots of private fine-tuning.
@tunguz
Bojan Tunguz
1 year
What I would *REALLY* love to have is a private version of ChatGPT that’s been trained on your internal org documents. Have an HR question? Ask Chat. Don’t know how to use some application? Ask Chat. When is the next vacation day? Ask Chat.
535
455
8K
0
0
7
@NathanThinks
nathan lile
9 years
I'm running for write-in student body president
0
2
6
@NathanThinks
nathan lile
1 year
Quasi-quilling "The Dark Side of a Starry Night" 🌚 #AIart 🤝 Van Gogh 🤝 Pink Floyd 🤝 using #DeepFloydIF #DeepFloyd from the @deepfloydai band
Tweet media one
Tweet media two
Tweet media three
Tweet media four
@NathanThinks
nathan lile
1 year
Quilling The Dark Side of the Moon 🌚 requested by @danberridge 😊 #DeepFloydIF #DeepFloyd @deepfloydai
Tweet media one
Tweet media two
Tweet media three
Tweet media four
4
2
16
0
1
6
@NathanThinks
nathan lile
9 years
See through awards are not cool when it comes to hand placement #LOTSMC15 @McKellogs @drjosephmoore http://t.co/YqZB67HBtR
Tweet media one
1
1
6
@NathanThinks
nathan lile
4 months
Current approaches for alignment, evaluation, and dynamic adaptation of AI models have shortcomings we aim to redefine through cutting-edge research Check out the great article by  @rachelmetz @technology
0
0
7