Thomas H. Chapin IV @tomchapin Twitter profile

Pinned Tweet

Thomas H. Chapin IV

19 days

The “Healthspan, Longevity, and AI” community has been growing lately! If you are interested in the latest news about AI-related medical advances, feel free to join and contribute!

1

0

17

Last Seen Profiles

@amerlondon

@MMmmichis

@au_PAY_market

@Transgportraits

@W1H5K

@Asianplaygirluk

@therealsupdamo1

@hkesmfwp3l9tzm

@pepefaurio

@stw_pdg

@RUDESLLEIDA

@jeff_poor

@Marasbeyoglu

@LouiseOnAir

@Manolo_mma

@phreesia

@stefanoasr

@BTWsanctuary

@costinoncescu

@Hundilisse

@pawsvtuber

@ParksTayousei

@shrimpjulia

@context_girl

@alinalal_

@ParksRecords

@DavidMelling1

@amquilli

@Mason_Simpson11

@brenotsaucy

@gemmanofficial

@SacBIRMINGHAM

@RColville

@chiba_yuto_bb

@emyuko

@NewDealChief

Thomas H. Chapin IV

@tomchapin

6 months

OpenAI is quite clever. By having people independently create "GPTs", each with their own collection of uploaded knowledge and data, OpenAI is able to get around all of the pesky copyright restrictions that are currently holding back ChatGPT.

115

283

3K

Thomas H. Chapin IV

@tomchapin

6 months

So I learned something new today. I was frustrated with ChatGPT because it kept interrupting me, but apparently, all I have to do is press and hold down the circle button in the middle of the screen while I am talking. I just assumed that it was a hands-free interface. I feel

165

83

1K

Thomas H. Chapin IV

@tomchapin

26 days

Would you trust an Ai-powered humanoid robot in your home if its code was not fully open source and auditable? I wouldn’t.

Tsarathustra

@tsarnick

26 days

Jensen Huang says in the near future, humanoid robots will be available for $10-20,000

93

210

1K

322

74

972

Thomas H. Chapin IV

@tomchapin

6 months

This is too good not to share. A local AI model running on a MacBook is now capable of describing what it sees via a live web cam feed. Something like this could give “sight” to vision impaired people! What other applications can you think of? Robots? AI Agents with vision?

Robert Lukoszko — e/acc

@Karmedge

6 months

So I set up BakLLaVA-1 in the llama.cpp, and now it can provide real-time descriptions of the live feed from my camera check it out! open source? cc: @nisten @thursdai_pod @willdepue #llama

62

127

866

9

53

440

Thomas H. Chapin IV

@tomchapin

11 months

@RyzeFi @MrBeast Stop trying to make the mark of the beast happen

10

1

378

Thomas H. Chapin IV

@tomchapin

5 months

GPT-4, when prompted with this new “MedPrompt” technique, outperforms even Med-PaLM 2 With Medprompt, GPT-4 achieves state-of-the-art results on all nine of the benchmark datasets in the MultiMedQA suite. The method outperforms leading specialist models

Steering at the Frontier: Extending the Power of Prompting - Microsoft Research

We’re seeing exciting capabilities of frontier foundation models, including intriguing powers of abstraction, generalization, and composition across numerous areas of knowledge and expertise. Even...

www.microsoft.com

3

0

342

Thomas H. Chapin IV

@tomchapin

4 months

@TimSweeneyEpic @alisabets Next thing you know, Apple will start requiring a cut of every online purchase that is made while using Safari. Are they taking 27% of every purchase that someone makes via the Amazon app?

7

19

309

Thomas H. Chapin IV

@tomchapin

3 months

I lost a ton of respect for @lexfridman when I saw him block so many people on Twitter. I also lost a ton of respect for @NPCollapse when he blocked @BasedBeffJezos after their public debate. If you truly love people, you don’t block them. You listen and tolerate them, while

92

10

300

Thomas H. Chapin IV

@tomchapin

26 days

@Drachs1978 I always thought it was because he was scared of being sued by Nintendo for copying the green pipes from Mario. He never outright said it, but I remember a bunch of noise on social media about it starting to reach a crescendo right when he took it down. It’s a real shame

7

5

281

Thomas H. Chapin IV

@tomchapin

1 month

Do you remember being born? I do. I remember even before I was born. I remember floating in the darkness, hearing my mother’s heartbeat, wondering what this all was. I remember the intense pressure on my entire body as I was born. I remember the first time I saw light. I

175

9

285

Thomas H. Chapin IV

@tomchapin

3 months

Just got back home after another intense hackathon weekend for the “01” project. The team is now on week 4 of the mission to build an open source version of the Rabbit r1, and is making steady progress! Git repo:

21

17

258

Thomas H. Chapin IV

@tomchapin

3 months

My afternoon - A foggy day in Seattle, hanging out with insanely talented AI people at a hackathon to build an open source Rabbit R1 alternative. The project is called 01, and the group is now on the third week of the mission. Git repo:

12

18

248

Thomas H. Chapin IV

@tomchapin

1 month

is absolutely Incredible.

45

12

249

Thomas H. Chapin IV

@tomchapin

2 months

I am officially giving Anthropic $20 a month for Claude 3 Opus, while also giving OpenAI $20 a month for ChatGPT Plus. Both models have particular strengths and weaknesses, and both are extremely impressive. Would you pay for access to both, or would you only pick one?

129

11

241

Thomas H. Chapin IV

@tomchapin

3 months

Jensen Huang (CEO of Nvidia) says that we should be using *engineering* to solve the complex problems in biology and life science.

12

34

205

Thomas H. Chapin IV

@tomchapin

11 months

I am currently looking for work! If you or your company is tackling the problem of human aging/longevity and could use some extra manpower, please message me! I’m eager to apply my 25+ years worth of programming skills and knowledge about AI to accelerate your progress! If I

5

35

176

Thomas H. Chapin IV

@tomchapin

6 days

Real talk. What do you think about the idea of an AI agent (that can reliably perform valuable work for money), that has been tasked with the purpose of earning and evenly distributing an AI-funded universal dividend to all living humans? Pros? Cons?

104

15

187

Thomas H. Chapin IV

@tomchapin

8 months

I taught my 7 year old how to ask questions to ChatGPT on his iPad (via the voice recognition feature), then I showed him how to prompt ChatGPT to respond at his reading level. He has spent the last hour asking numerous questions about his favorite TV shows and movies, and

24

22

179

Thomas H. Chapin IV

@tomchapin

9 months

@lisatomic5 Red: 100% of people have to pick it, for 100% of people to live. Any less than 100% of people picking it results in a LOT of death. Even if 99% of people pick it, a ton of people end up dying. Blue: Only 50% of people have to pick it, for 100% of people to live. If you have

11

2

172

Thomas H. Chapin IV

@tomchapin

10 months

Should I attempt a 31 Day Trial Run of @bryan_johnson 's Blueprint Diet? *The Green Giant*: - 20 oz of water - Spermidine, 2 Tbsp chlorella powder, yielding 13.5mg spermidine* - Amino Complex 7.6 g - Creatine 2.5 g - Collagen Peptides 20 g - Cocoa Flavanols 500 mg - Ceylon

27

5

148

Thomas H. Chapin IV

@tomchapin

18 days

As someone who works with AI and voice tech, I have often wondered how many of the people on X are actually real. It’s not hard to create a fake profile or to simulate a voice. I follow a lot of people that I suspect are fake. We need anonymous proofs of personhood.

57

4

149

Thomas H. Chapin IV

@tomchapin

1 month

Why does it seem like the only people making laws about AI are the people who want to decelerate things? Where is the e/acc lobbying group? Where are the law makers who are standing up and protecting ideas like open source, decentralized AI?

36

18

135

Thomas H. Chapin IV

@tomchapin

7 months

@d_feldman Ah, yes

1

3

143

Thomas H. Chapin IV

@tomchapin

9 months

I want an AI tool that follows all of my feeds, my discord groups, Twitter/X, Reddit, Facebook, the news, my email, etc, and I want it to organize, rank, and summarize the most important information for me every day, based off of my life goals and personal interests

26

5

136

Thomas H. Chapin IV

@tomchapin

1 month

Accepting aging and death as unavoidable is incredibly short-sighted. If we spent even a small fraction of our GDP on aging research and on lifting of those less fortunate than ourselves (instead of on things like war and death and destruction), we might actually have a chance.

໊

@buffys

1 month

what opinion will you defend like this?

1K

156

2K

13

17

137

Thomas H. Chapin IV

@tomchapin

11 months

If a magical medical treatment existed that would reverse your physical age and health to back when you were in your prime, would you take it, or would you try to have it outlawed?

I would take it

498

I would have it outlawed

81

Other (explain)

47

40

21

111

Thomas H. Chapin IV

@tomchapin

4 months

@Nexuist Check out this video from Tech Ingredients, too! The guy is like a real life “Doc” Brown from Back to the Future. He is always building crazy gadgets. This time he made an AI powered laser that destroys moving targets:

We built an AI directed laser that destroys moving targets...

We built an AI laser. Unlike our other rocket motors, this one can be controlled.Find us on Patreon our website and twitter/x:https://www.patreon.com/teching...

www.youtube.com

2

15

125

Thomas H. Chapin IV

@tomchapin

4 months

Extension of healthy life. Do you think this is a worthy cause?

Yes

197

No

40

Just show me the results

84

4

11

115

Thomas H. Chapin IV

@tomchapin

3 months

For an AI model to produce coherent long-form video like this, it has to have a visual understanding of physics and the entire world encoded inside of it. AI can now dream synthetically. This might be bigger than GPT-5.

OpenAI

@OpenAI

3 months

Introducing Sora, our text-to-video model. Sora can create videos of up to 60 seconds featuring highly detailed scenes, complex camera motion, and multiple characters with vibrant emotions. Prompt: “Beautiful, snowy

10K

33K

141K

21

12

112

Thomas H. Chapin IV

@tomchapin

4 months

“Biologists discovered that mitochondria in different tissues talk to each other to repair injured cells. When their signal fails, the biological clock starts winding down.” “…genetically damaging the mitochondria extended the worms’ lives by 50%. That

Cells Across the Body Talk to Each Other About Aging | Quanta Magazine

Biologists discovered that mitochondria in different tissues talk to each other to repair injured cells. When their signal fails, the biological clock starts winding down.

www.quantamagazine.org

1

110

Thomas H. Chapin IV

@tomchapin

6 months

If we had UBI (or some kind of Universal Dividend from AI profits), I bet humanity could solve aging. People would actually be able to dedicate more time to education and research, and on meaningful problems, instead of working jobs they hate just to put food on the table.

19

14

103

Thomas H. Chapin IV

@tomchapin

8 months

Very interesting to hear @elonmusk talk about the ideal future with AI, universal basic (or high) income, and the concept of death as a choice. Sounds like he finally gets it!

Thomas H. Chapin IV

@tomchapin

9 months

It’s not about being forced to live forever. Instead, it’s about having the option of living *for as long as you want*.

14

3

84

17

20

100

Thomas H. Chapin IV

@tomchapin

1 year

Scribble in real time with Stable Diffusion Found at: GitHub repo:

1

8

104

Thomas H. Chapin IV

@tomchapin

6 months

@ZekeGeiss “It’s totally hopeless to compete with us”

4

8

103

Thomas H. Chapin IV

@tomchapin

7 months

ChatGPT’s voice chat is pretty amazing, but the conversation still doesn’t feel natural. If I pause for even a moment, it assumes I am done talking. Has anyone tried training an AI model with the sole purpose of determining when someone has finished talking and is ready for a

12

4

89

Thomas H. Chapin IV

@tomchapin

10 months

Creating a digital twin of yourself is NOT immortality

35

7

83

Thomas H. Chapin IV

@tomchapin

9 months

It’s not about being forced to live forever. Instead, it’s about having the option of living *for as long as you want*.

Elon Musk

@elonmusk

9 months

“May you live forever” is the worst possible curse once you understand deep time

25K

27K

260K

14

3

84

Thomas H. Chapin IV

@tomchapin

6 days

Everyone is wondering what OpenAI’s announcement will be about on Monday. Maybe related to their “Voice Engine” trademark that was filed back in March?

11

4

87

Thomas H. Chapin IV

@tomchapin

9 months

@RokoMijic @lisatomic5 The key is to understand not what is the “correct” answer, but the answer that will MOST LIKELY be chosen by the other people that you don’t have control over. The poll utterly proves this. Sorry, but blue is the best choice to vote for.

3

0

79

Thomas H. Chapin IV

@tomchapin

6 months

An Aging Expert Thinks Humans Can Live for 20,000 Years. He's Not Crazy. A recent “Popular Mechanics” article featuring @jpsenescence “He believes that the medical intervention necessary for these extremely long lives will require finding a way to

An Aging Expert Thinks Humans Can Live for 20,000 Years. He's Not Crazy.

There's one major key to unlocking extreme longevity.

www.popularmechanics.com

7

9

77

Thomas H. Chapin IV

@tomchapin

4 months

A lot of my opinions on unregulated capitalism are formed from the years that I personally manipulated “free” markets on various MMORPG games (such as World of Warcraft). I hit gold cap in those games by using my capital in the game to control and manipulate the open markets

21

4

76

Thomas H. Chapin IV

@tomchapin

5 months

To say that death is inevitable and unavoidable is to deny the possibility that it might not be. This is the essence of deathism, and the opposite of existential hope. Deathism is the denial of first principles thinking. It is a denial of life’s ability to discover and conquer

18

5

74

Thomas H. Chapin IV

@tomchapin

6 months

A loaf of bread at the grocery store is now almost $5. A 5 lb bag of flour is $2.50 Yeast packets are also around 50 cents, or you can buy in bulk. All you need for bread is flour, water, yeast, and a little bit of salt. Math: A loaf of bread costs $1 if you bake it yourself!

38

6

73

Thomas H. Chapin IV

@tomchapin

9 months

A great explanation of 3D Gaussian Splatting, a relatively new technology that can turn photos and videos into 3D, photorealistic environments that you can dynamically navigate. VR video games that use this technology in the future are going to be absolutely insane.

dylan

@dylan_ebert_

9 months

a two minute explanation of 3D Gaussian Splatting

48

384

2K

5

14

73

Thomas H. Chapin IV

@tomchapin

7 months

Checking out LM Studio, and I have to say that I'm pretty impressed! It makes it really easy to download open source language models from HuggingFace and get them up and running in a chat interface. It actually supports both Mac *and* Windows, too!

5

7

74

Thomas H. Chapin IV

@tomchapin

13 days

I wonder how long it will be until we can wear a Neuralink while we are sleeping, and automatically turn our dreams into videos with an AI video generation model like Sora?

26

9

69

Thomas H. Chapin IV

@tomchapin

1 month

It really makes me sad how the X algorithm penalizes me for following more people than follow me. There are so many cool people in the AI and longevity space that I run across and want to follow, but I have to wait until my follower count grows.

12

0

69

Thomas H. Chapin IV

@tomchapin

8 months

What if we didn’t *have* to die? Imagine a future (say 500-1,000 years) where we have used technology such as AI to conquer disease and the process of aging itself. Most people I surveyed agree that this future is inevitable. Knowing this, why are we content to let the “boat

15

5

65

Thomas H. Chapin IV

@tomchapin

9 months

What happens when your job is automated by an AI service that costs a tiny fraction of your salary, and your boss lays you off? What happens when you can’t find another job because your entire 20 year career path has been nullified by AI automation? What kind of job will you do

Scott Santens

@scottsantens

9 months

Time to distribute an AI dividend to make sure people can buy what AI creates.

18

140

459

13

12

65

Thomas H. Chapin IV

@tomchapin

6 months

To clarify: I’m not saying their cleverness here is a bad thing. I’m a big fan of OpenAI! OpenAI is trying to play by the rules, but still give people the tools they want. Their end goal is to benefit all of humanity with safe and benevolent AGI. They have been doing their best

5

3

64

Thomas H. Chapin IV

@tomchapin

3 months

In case anyone is curious, I returned my Apple Vision Pro today. They have built an amazing product, but it is seriously lacking in a number of areas. I’m really looking forward to seeing how they address these issues in the next version: 1. Too heavy/uncomfortable (no matter

Sheel Mohnot

@pitdesi

3 months

Zuck on the Apple Vision Pro He's still got fire in him, love to see it

2K

6K

74K

11

6

64

Thomas H. Chapin IV

@tomchapin

1 month

Reminder: Don’t pre-pay for an entire annual subscription for *any* AI-based online service. Things are changing too fast for that.

6

2

65

Thomas H. Chapin IV

@tomchapin

4 months

I canceled my X premium+ account because Grok isn’t useful for anything that I want to use it for, and X was still showing me ads all over the place (albeit not in my main feed). @elonmusk Please give me a version of X with *zero* ads. PLEASE.

8

3

65

Thomas H. Chapin IV

@tomchapin

8 months

Hey e/acc people: The only “banger” tweets I care about are ones that actually provide value. Stop yammering, and show me the cool AI tools you are working on. Especially ones related to health and longevity!

13

1

57

Thomas H. Chapin IV

@tomchapin

7 months

Self-Taught Optimizer (STOP) - A new AI programming method from Microsoft AI processes that recursively self improve their own code. They released the paper (which includes code samples), but haven’t released open source code repo on GitHub yet? Arxiv Paper:

6

5

63

Thomas H. Chapin IV

@tomchapin

7 months

I asked AI to imagine someone surfing web pages, and it produced this image:

6

1

62

Thomas H. Chapin IV

@tomchapin

5 months

@LinusEkenstam It’s not real until it’s open source and locally runnable with encrypted secrets. Anyone who hands over their entire identity to a cloud provider is a…

3

1

61

Thomas H. Chapin IV

@tomchapin

9 months

@jxnlco Sounds like we need a service that can actually extrapolate, augment, and refine a training dataset, when given a very small example of the over-arching goal for the model?

5

0

60

Thomas H. Chapin IV

@tomchapin

7 months

Whoa! I was just talking the other day about how AI voice recognition needs to be able to understand nuanced emotion and emphasis, and then this releases: CLARA (Multilingual Contrastive Learning for Audio Representation Acquisition) – a project designed

CLARA: Advancing Machines in Understanding Speech Nuances | LAION

Voices carry not only words but also convey emotions, emphasis, and nuance through aspects like tone and accent. However, existing speech technology only ...

laion.ai

1

13

60

Thomas H. Chapin IV

@tomchapin

8 months

They did it! I knew it was coming soon, but didn’t expect it this fast! Realistic 3D videos in VR that are fully navigable and viewable from any angle, made possible with Gaussian splatting!

Andrey Volodin

@s1ddok

8 months

I imagine that inventing a TV felt somewhat similar. 🤯 Gaussian splatting video in VR, now running at 2x2224x2160 @ 100 FPS. iPhone's Spatial Videos simply can't compare, temporal flickering is there to remind your brain you're in VR! 🧵on performance, details, and next steps:

42

160

1K

4

12

57

Thomas H. Chapin IV

@tomchapin

8 months

The “Healthspan, Longevity, and AI” community has been growing lately! If you are interested in the latest news about AI-related medical advances, feel free to join and contribute!

6

4

53

Thomas H. Chapin IV

@tomchapin

6 months

Humanity needs to evolve past this deathist way of thinking. A candle’s meaning comes from the light it gives. It would be stupid to snuff out a candle just because the darkness would make you appreciate the light more.

Robert Greene

@RobertGreene

6 months

Life has more meaning in the face of Death.

97

647

4K

8

6

55

Thomas H. Chapin IV

@tomchapin

9 months

“By programming cells to constantly switch between two aging pathways, researchers were able to prevent them from fully committing to either deteriorative process, a method that nearly doubled the lifespan of the cells.”

Scientists Dramatically Extend Cell Lifespan in Anti-Aging Breakthrough

“I don’t see why it cannot be applied to more complex organisms,” said the study's senior author.

www.vice.com

1

6

55

Thomas H. Chapin IV

@tomchapin

4 months

All these people making 2024 predictions, while forgetting the Apple Vision Pro is about to launch. Imagine conversational AI that can see what you see, but in real time. 2024 is going to be a crazy year.

5

4

57

Thomas H. Chapin IV

@tomchapin

8 days

Elon Musk shares his opinions on the future of education, and how to make it more engaging with interactive and relevant learning experiences.

9

12

56

Thomas H. Chapin IV

@tomchapin

1 month

The first step to obtaining human longevity is spreading the dream that such a thing is even possible.

8

5

56

Thomas H. Chapin IV

@tomchapin

7 months

Think RAG (retrieval augmented generation), but instead of just querying vector databases, you are querying small models that are fine tuned on specific domains of knowledge. Now build a system to fine tune and iterate a huge hive of these smaller “expert” models automatically.

11

6

55

Thomas H. Chapin IV

@tomchapin

6 months

Someone leaked news about OpenAI’s new GPT-4-Turbo model, which supposedly has a 128k token limit, and is a third the price of GPT-4. Never mind entire books! This will give people the ability to feed entire *codebases* into the prompt! Also, on the open source side of things,

11

2

51

Thomas H. Chapin IV

@tomchapin

26 days

@BobbyGRG The difference here is that a humanoid robot can do things like carry a sharp kitchen knife around the house while my family is asleep. You are willing to put that amount of trust into a closed source system that is entirely owned and controlled by a for-profit corporation?

20

0

53

Thomas H. Chapin IV

@tomchapin

2 months

Very proud to be a part of this! You can be a part of it, too. Come help us build 01! All you have to do is fork the repo, complete a feature, and submit a pull request!

GitHub - OpenInterpreter/01: The open-source language model computer

The open-source language model computer. Contribute to OpenInterpreter/01 development by creating an account on GitHub.

github.com

Open Interpreter

@OpenInterpreter

2 months

100 years in the making. 100 hours to go.

25

38

466

1

4

52

Thomas H. Chapin IV

@tomchapin

11 months

What if someone built a decentralized mesh version of Twitter that allowed people to import all of their tweets (as well as the tweets from the people they follow)?

6

9

48

Thomas H. Chapin IV

@tomchapin

8 months

It’s a real shame that there are so few modern movies or TV shows that show AI and technology being used in an actual *positive* way to improve life for humanity. Movies almost always depict dystopian scenarios where everything has gone wrong. What if things WENT RIGHT?

8

4

47

Thomas H. Chapin IV

@tomchapin

4 months

Nabla is building an AI assistant for doctors that automatically writes clinical notes

Nabla raises another $24 million for its AI assistant for doctors that automatically writes...

Paris-based startup Nabla just announced that it has raised a $24 million Series B funding round for its AI copilot for doctors.

techcrunch.com

2

0

49

Thomas H. Chapin IV

@tomchapin

8 months

Just finished watching this awesome video (28 min) about new life extension and anti-aging discoveries, and I put together a summary. A recent breakthrough in aging research has identified key players in the aging process: senescent cells. These are

12

5

44

Thomas H. Chapin IV

@tomchapin

2 months

Nvidia announces AI-powered health care 'agents' that outperform nurses - and cost $9 an hour Hippocratic AI and Nvidia teamed up to develop an 'empathetic' health care bot that handles patient calls More info:

Nvidia announces AI-powered health care 'agents' that outperform nurses — and cost $9 an hour

Nvidia has partnered with artificial intelligence health care company Hippocratic AI to develop an "agent" that outperforms nurses on phone calls with patients.

www.foxbusiness.com

2

1

45

Thomas H. Chapin IV

@tomchapin

9 months

Does anyone know of any good conversational, AI powered real estate tools out there? I want to be able to just describe the type of property that I’m looking for (as if I’m talking with a real estate agent), and have results instantly returned. For example: “Please find the

18

2

45

Thomas H. Chapin IV

@tomchapin

8 months

I love how @aubreydegrey brought the concepts of healthspan, longevity, and “Say Forever” together in perfect harmony

Say Forever!

@SayForeverOrg

8 months

Side affect we’re all in favor of! @aubreydegrey is saying forever at @raadfest #sayforever #raadfest2023

4

21

89

4

5

44

Thomas H. Chapin IV

@tomchapin

6 months

@iamdiegopy I agree with you completely, and have been pondering this problem and how it might be solved for some time now For example:

Thomas H. Chapin IV

@tomchapin

7 months

ChatGPT’s voice chat is pretty amazing, but the conversation still doesn’t feel natural. If I pause for even a moment, it assumes I am done talking. Has anyone tried training an AI model with the sole purpose of determining when someone has finished talking and is ready for a

12

4

89

7

1

44

Thomas H. Chapin IV

@tomchapin

6 months

Recruiter on LinkedIn: “Can you send me your resume?” “Sure! Let me just print out my LinkedIn profile and fax it on over to you. Or should I send it via pony express?”

8

0

43

Thomas H. Chapin IV

@tomchapin

3 months

My dream of an AI system that will let me create an inventory of everything in my house (just by looking around) is slowly but surely becoming a reality

Ryan Carson

@ryancarson

3 months

If you have compute hungry AI ML tasks, you can spin up an Intel Gaudi2 instance and watch it just rip. I wrote a quick Python script to use the facebook/detr-resnet-50 model (🙏🏻 @huggingface ) to detect and label objects in a video. It's interesting that no one is talking

16

33

195

9

4

43

Thomas H. Chapin IV

@tomchapin

6 months

Another AI agent swarm architecture. Building agents without OpenAI? Video Link: Code from video: CoALa: MemGPT: TutorGPT: MetaGPT:

0

4

42

Thomas H. Chapin IV

@tomchapin

2 months

Idea: An AI system that submits my question to all of the top LLMs (GPT-4, Claude 3 Opus, Gemini, etc), and then uses an AI model with a large context window to compile them all into a single response (with only the best information from each of the different responses).

19

1

40

Thomas H. Chapin IV

@tomchapin

1 month

@iamethanglenny I’m not sure what it means, exactly? I have had at least one other person direct message and tell me that they remember being born, too, so that was super cool. I think there is a whole lot about human memory and the human experience that we still don’t fully understand.

3

0

42

Thomas H. Chapin IV

@tomchapin

9 months

@jerryjliu0 @md_rumpf One trick you can use is to take each chunk (along with a summary of the whole), and ask an LLM things like, “what questions does this chunk of text possibly answer?”, and have it generate question and answer pairs that link back to the larger chunk. You can basically build a

4

2

41

Thomas H. Chapin IV

@tomchapin

8 months

Your daily reminder to work on the problem of aging

Casey Handmer, PhD

@CJHandmer

8 months

If we don't solve aging we're all going to die.

94

100

1K

3

38

Thomas H. Chapin IV

@tomchapin

7 months

Earlier this evening, I showed my 7 year old kid the new image generation features in ChatGPT, and he was giddy with excitement. Over the next hour, he used the voice recognition feature to describe and generate all sorts of crazy pictures. Here are a few of them.

9

2

38

Thomas H. Chapin IV

@tomchapin

6 months

David Shapiro announces construction of an autonomous AI agent swarm: “It's time to knuckle down and get serious. We are SO CLOSE to fully autonomous AGI. I'm telling you guys. The last four years of my research and work has been building up and preparing for this moment. In

5

2

39

Thomas H. Chapin IV

@tomchapin

8 months

All of the white collar workers using Windows 11 Copilot are being recorded, and the data will be used to train their AI replacements, right? I’m really curious what the world is going to look like in 5 years… What is going to happen to all of the people without jobs?

19

1

39

Thomas H. Chapin IV

@tomchapin

6 months

I agree with all 5 points of The Vitalist Declaration. How about you?

Adam Gries

@adamgries

6 months

Vitalism (organized as a 501c4 non profit) is meant to be the magnet to bring together those who will drive the revolution. Anybody can be a Vitalist by embracing the Vitalist Declaration:

1

8

53

8

2

37

Thomas H. Chapin IV

@tomchapin

10 months

@ilyasut I have ingested this tweet, and am now superior to people who haven’t read it yet!

0

36

Thomas H. Chapin IV

@tomchapin

5 months

One thing 2023 has taught me is that we can shape the future by interacting with others on X. Even liking a post can have a surprising impact. If only 1 out of 100 people who view a post actually like it, then our likes are quite rare and valuable. Let's pay attention to the

5

4

37

Thomas H. Chapin IV

@tomchapin

2 months

Finally got access to Gemini 1.5 Pro with the 1 million token context window! Hmm… What should I try out on it first? Maybe a big codebase?

9

0

37

Thomas H. Chapin IV

@tomchapin

1 month

@jenny____ai Sounds like that should be a work-paid phone to me?

0

36

Thomas H. Chapin IV

@tomchapin

8 months

Side note: The way how he prompts ChatGPT to respond at his reading level is simply by verbally telling it something like, “please respond with words that a seven year old could read”, or by asking it to “please use smaller words that are a little bit easier for me to read”.

3

37

Thomas H. Chapin IV

@tomchapin

8 months

- Meta announces AI camera glasses. - Tab announces their AI pendant. - Humane announces their AI pin. - Rewind announces their AI pendant. Consumer wearables that record everything you do and process it with AI are the hottest new thing!

Dan Siroker

@dsiroker

8 months

Introducing Rewind Pendant - a wearable that captures what you say and hear in the real world! ⏪ Rewind powered by truly everything you’ve seen, said, or heard 🤖 Summarize and ask any question using AI 🔒 Private by design Learn more & preorder:

355

193

1K

10

3

35

Thomas H. Chapin IV

@tomchapin

1 month

@AnkurPandey You, and everyone else I’ve ever met! ☺️

4

0

36

Thomas H. Chapin IV

@tomchapin

4 months

@Jason It’s not about socialism or capitalism. It’s about balance. Capitalism, balanced by elements of socialism. Why do so many people insist that the only option is one extreme vs the other?

23

2

36

Thomas H. Chapin IV

@tomchapin

6 months

I’ve been noticing a number of AI decentralization projects lately. Just off the top of my head: - Petals - Shoggoth - GPUnet - Arbius What are some other ones that I missed?

Aaron Ng

@localghost

6 months

> Year 2003: Seeding torrents for decentralized file sharing. > Year 2023: Hosting an LLM for decentralized intelligence sharing.

18

35

358

9

3

34

Thomas H. Chapin IV

@tomchapin

24 days

OpenCRISPR - Open source gene editing with AI

Ali Madani

@thisismadani

24 days

Can AI rewrite our human genome? ⌨️🧬 Today, we announce the successful editing of DNA in human cells with gene editors fully designed with AI. Not only that, we've decided to freely release the molecules under the @ProfluentBio OpenCRISPR initiative. Lots to unpack👇

282

1K

4K

0

34

Thomas H. Chapin IV

@tomchapin

29 days

@summeroff And yet now we have AI, which we, the lesser beings, are creating. But are we really creating it, or are we just discovering the inevitable and emergent intelligence of the Universe itself?

3

1

34