Simon Suo @disiok Twitter profile | Pikagi

Pikagi

Simon Suo

@disiok

1,932

Followers

1,253

Following

59

Media

894

Statuses

co-founder @llama_index prev: AI research @Waabi_ai @Uber_ATG , PhD AI @UofT

San Francisco, CA

https://t.co/vFZ3PJfRPJ

Joined June 2010

Don't wanna be here? Send us removal request.

Pinned Tweet

@disiok

Simon Suo

11 months

After 5 amazing years at @Waabi_ai & @Uber ATG, surrounded by the most brilliant minds in AI and mentored by @RaquelUrtasun , I decided to embark on a new adventure with my close friend @jerryjliu0 . Today I’m excited to announce that we have started a company around @llama_index …

@jerryjliu0

Jerry Liu

11 months

I’m super excited to make it official: @disiok and I have started a company around @llama_index , and we’ve raised a $8.5M seed round led by @GreylockVC ! 🔥🚀 We are building the open-source data framework to unlock LLM capabilities on your private data.

96

120

1K

5

9

61

Last Seen Profiles

@INTERNETC0RE

@Callme_kendrick

@MayoBooks

@xjr_69r

@siyemmm

@engrmalam635

@DimasPivass

@emmagreenmodel

@IAMRockwell

@Kelly3770387531

@transmistress00

@Woking_Cons

@jesskettle

@yakorunasapo

@RuthRuth1228247

@thebigword

@alexbelluwtsd

@akuneko_shop

@CrystalLustxoxo

@AlamResalat

@04tshc

@BGS_Trips

@DAZNTaiwan

@erwitami

@PpaAcademy

@nadyanarita

@nickmilaniXXX

@cmsoken

@shimpufishy

@_zhnec

@michael_nc44363

@9990lol_09

@laysthailand

@JMafoutaOff

@patryjordan

@nolankromke

@disiok

Simon Suo

10 months

Llama 2 Prompt Structure is a fantastic playground for the latest Llama 2 model. But wait, is the suggested prompt structure the best way to interact with the model? """ User: <your prompt goes here> Assistant: """ Some initial notes 👇

Tweet card media

meta/llama-2-13b-chat – Replicate

A 13 billion parameter language model from Meta, fine tuned for chat completions

2

34

164

@disiok

Simon Suo

4 months

2023 is a whole lot of “let’s do it” - startup - move to SF - hire team - new relationship - mindset upgrades Grateful for family, friends, believers, advisors. The journey is just beginning.

Tweet media one

7

6

97

@disiok

Simon Suo

10 months

Llama 2 Protip If you are hacking with the @replicate hosted 13B or 70B models (playground or API), this is relevant for you! The `max_length` parameter limits the total number of tokens (i.e. sum of input prompt + output response). More details 👇

3

8

70

@disiok

Simon Suo

18 days

How I visualized the so-very-overloaded AI engineer role (a la @swyx 's ) anyone aware of breakdown / articulation similar to this?

Tweet media one

8

4

59

@disiok

Simon Suo

8 months

prototype: one weekend productionize: weeks? months? We are tackling this gap by open-sourcing production-quality app templates, so you can hit the ground running. goes way beyond the typical streamlit demo, complete with advanced features like -…

@llama_index

LlamaIndex 🦙

8 months

We’re excited to open-source - a full-stack, production-ready RAG app! 🦙🏦 Supports streaming, reasoning steps, citations, intuitive UI This can save you weeks/months of hard work in trying to build a prod LLM app from scratch🔥

10

102

567

1

1

46

@disiok

Simon Suo

3 years

It's a nudge! It's a U-turn! It's S̶u̶p̶e̶r̶m̶a̶n̶ TrafficSim! Excited to share our work on learning to simulate realistic multi-agent behaviors at #CVPR2021 . Come to our poster session at 10pm ET today to learn more! Joint work with Sebastian, @sergioksas , and @RaquelUrtasun

3

7

43

@disiok

Simon Suo

4 years

Check out at #ECCV2020 how we generate multiple scene-consistent motion forecasts of complex urban traffic, directly from sensor data! Paper: Live Q&A: Thu 9am & 7pm ET Joint work with: @sergioksas , @cole_gulino , @katielulula , @lrjconan , @RaquelUrtasun

1

14

40

@disiok

Simon Suo

9 months

Exciting day! @OpenAI GPT-3.5 Turbo fine-tuning is finally out!

GPT-3.5 Turbo fine-tuning and API updates

Developers can now bring their own data to customize GPT-3.5 Turbo for their use cases.

0

4

35

@disiok

Simon Suo

9 months

My notes on the new @OpenAI fine-tuning release: 1. currently limited to GPT-3.5 Turbo with 4K token (GPT-4, function-calling, and gpt-3.5-turbo-16k all coming this fall) 2. similar DX as old GPT-3 fine-tuning API, just list of chat messages instead of (prompt, completion) pairs…

Tweet media one

Tweet media two

@disiok

Simon Suo

9 months

Exciting day! @OpenAI GPT-3.5 Turbo fine-tuning is finally out!

0

4

35

5

6

34

@disiok

Simon Suo

1 year

Similar to autonomous driving, maybe we should be explicit about automation levels for knowledge work as well. L5: fully autonomous agents (AutoGPT et al.) L4: semi-autonomous with human-in-the-loop support L3: copilot, assistant, constrained cognitive tasks

@ylecun

Yann LeCun

1 year

Haha. Auto-Regressive LLMs gonna auto-regress. Your hands must remain on the keyboard at all time. Level-2 Writing assistance? Yes! Level-5 autonomous writing? No! "Here’s What Happens When Your Lawyer Uses ChatGPT"

60

96

673

1

5

32

@disiok

Simon Suo

2 months

You can build an AI consulting shop by just always replying “look at your data more”

3

1

31

@disiok

Simon Suo

11 months

Designing a robust system requires clear interfaces and well-behaved components. The future of LLM-powered systems is not one "monolithic agent" that does everything. It will be many specialized components (think query routing, knowledge retrieving, API calling). New @OpenAI …

1

2

28

@disiok

Simon Suo

1 year

Watching @karpathy 's amazing talk and came across a familiar diagram

Tweet media one

1

2

25

@disiok

Simon Suo

1 year

Are your agents too dumb or you just need better tools?

Tweet card media

Dumber LLM Agents Need More Constraints and Better Tools

Summary

0

4

25

@disiok

Simon Suo

4 months

Libraries need to start writing docs/tutorials for LLMs instead of humans soon

@bernhardsson

Erik Bernhardsson

4 months

I’m bullish on LLMs for code but this is kind of a funny/weird counterpoint that maybe it makes developers chose worse tools

Tweet media one

85

48

734

5

3

24

@disiok

Simon Suo

11 months

If you are at #BerkeleyHackathon , definitely come check out @jerryjliu0 talk on @llama_index

Tweet media one

0

1

21

@disiok

Simon Suo

2 months

We are in the era where the metric is just vibes

@_aidan_clark_

Aidan Clark

2 months

how the hell did anthropic get away with this chart?

Tweet media one

18

2

234

3

0

20

@disiok

Simon Suo

3 months

@jxnlco

1

1

19

@disiok

Simon Suo

10 months

Few people know this: LlamaIndex got its name before the original Llama model came out. Great Minds Think Alike? 🤔 Incredible to see Meta continuing its commitment to open source. And glad we can stack more 🦙.

@llama_index

LlamaIndex 🦙

10 months

🦙x🦙 = 💪 Experiment with Llama 2 now via LlamaIndex! We made a special release (v0.7.10.post1) to help you get started super easily 👇

Tweet media one

1

16

112

1

2

18

@disiok

Simon Suo

4 months

one of the best writeup on assistant architecture in production (how did I miss this gem): 👏 @hiranya911

Tweet card media

A deep dive into the world's smartest email AI

A detailed look under the hood of how the Shortwave AI assistant works.

www.shortwave.com

1

5

17

@disiok

Simon Suo

4 months

get yourself some investors that keep your toes warm and keep you caffeinated thanks for the kindness and support @jerrychen @rischter_scale @GreylockVC 🙏

Tweet media one

Tweet media two

Tweet media three

1

1

17

@disiok

Simon Suo

11 months

@jerryjliu0 looking good

Tweet media one

0

1

17

@disiok

Simon Suo

3 months

RAG vs long context LLM? Why not both We just got new power to tune the desired point on accuracy vs latency/cost frontier. Exciting future ahead

2

2

16

@disiok

Simon Suo

6 months

im so confused is this a hard fork

1

0

16

@disiok

Simon Suo

10 months

Last but not least: using the reference encoding, seems like it works!

Tweet media one

Tweet media two

1

1

16

@disiok

Simon Suo

5 months

Do we really want AGI or just really cheap and fast open source thought calculators that runs in our laptop

2

3

16

@disiok

Simon Suo

6 months

X (twitter) ads revenue sharing + OpenAI GPTs revenue sharing = bracing myself for the sea of custom GPT tweets

2

3

15

@disiok

Simon Suo

5 months

Is reasoning inherently coupled with knowledge? My intuition says there’s a small kernel of knowledge that can bootstrap reasoning. Right now, to get improved reasoning, we also get useless facts. I just want my LLM OS to be pure compute, and not have the pre-installs.

6

1

15

@disiok

Simon Suo

2 months

I’m assembling a team of 10 IPhO kids. DM and you can get an immediate offer

7

1

15

@disiok

Simon Suo

5 months

next time I write a paper, I'm releasing it on torrent

1

2

13

@disiok

Simon Suo

19 days

Setting a goal to tweet an original / personal thought every day. Purely for the online to IRL friend pipeline. Optimizing for that tweet-to-friend ratio

6

0

14

@disiok

Simon Suo

10 months

A deep dive into the official codebase shows that system instructions and messages use specific formats (from ).

Tweet media one

Tweet media two

1

2

13

@disiok

Simon Suo

5 months

A small milestone on this wild ride! 🚀 Super grateful for the love from @llama_index community, and support from our investors @GreylockVC and partners. Excited for many more celebrations ahead with our incredible team.

@jerryjliu0

Jerry Liu

5 months

Excited to announce that @disiok and I are featured in @Forbes 30U30 2024! 🎉 It's been a wild year, and we couldn't have done this without our community, partners, investors, and of course our wonderful @llama_index team.

Tweet media one

35

21

316

0

1

13

@disiok

Simon Suo

1 year

Can we stop trying to replace lawyers and scientists with LLMs, let’s just figure out how to get really really good at analyzing multiple documents first…

2

0

13

@disiok

Simon Suo

11 months

ChatGPT Plugins have not reached PMF. This is unsurprising, but refreshing to hear it stated plainly by @sama . Browsing Redfin through ChatGPT (or any chat interface) is painful, I think for 2 reasons: 1. When my intent/goal is clear, I might as well use the website’s own…

@swyx

swyx @ICLR_conf

11 months

i also admire @sama 's ability to confront reality and state the truth - ChatGPT Plugins have not reached PMF! I don't wake up in the morning and book my travel and food thru ChatGPT. Instead, the apps are all integrating chat. The Western consumer continues to resist the…

Tweet media one

11

3

66

2

1

13

@disiok

Simon Suo

10 months

Feeling AGI @agihouse_org

Tweet media one

@jerryjliu0

Jerry Liu

10 months

If you’re around at the @agihouse_org hackathon today, come check out @disiok ’s talk on data agents! 🤖💾 - Advanced search/retrieval - Act over a large range of external APIs (email, search, etc. ) Get a head start with our full agent guide here:

0

5

32

1

3

13

@disiok

Simon Suo

3 years

It's been such an amazing journey! Can't wait to get started on the next chapter.

@RaquelUrtasun

Raquel Urtasun

3 years

Today is my last day @Uber after 4 wonderful years. Thank you @dkhos for your support and leadership. Solving #selfdriving is my life's passion and I'm super excited for what comes next. Will share an update soon. Stay tuned!

21

30

641

0

0

12

@disiok

Simon Suo

9 months

the world needs more thoughtful & comprehensive long-form content like this

@eugeneyan

Eugene Yan

9 months

Wrote abt patterns for LLM systems/products • Evals: Track performance • RAG: Add external knowledge • Finetuning: Improve specific tasks • Caching: Reduce latency & cost • Guardrails: Ensure output quality • Defensive UX: Anticipate & manage errors

38

180

801

0

1

12

@disiok

Simon Suo

4 years

Throwback to when I was a neural network for Halloween 😅

Tweet media one

0

0

12

@disiok

Simon Suo

3 months

When I think quality long form writing, @eugeneyan @lilianweng are the top 2 that come to mind

@eugeneyan

Eugene Yan

3 months

> For those trying to educate, consider writing longform, designed for someone to get "sweaty", especially in today's era of quantity > quality ... This is what I aspire to do. My audience will decrease. The ones that remain might not like it. But at least we'll learn something.

3

1

59

1

1

11

@disiok

Simon Suo

5 months

Is there an “uncanny valley” but for intelligence / reasoning

1

1

10

@disiok

Simon Suo

8 months

Evangelizing EDD (evaluation driven development)

@raydistributed

ray

@raydistributed

8 months

Want to learn how to build and evaluate production RAG app with @llama_index and @raydistributed ? Join #RaySummit Training Day! 1️⃣ Implement reliable eval methods for LLMs 2️⃣ Run experiments to optimize app components 3️⃣ Take best configs to production

0

7

18

0

0

10

@disiok

Simon Suo

11 months

AGI is gonna destroy us all . . . . Once it figured out how to consistently output some jsons

0

1

10

@disiok

Simon Suo

1 year

GPT4 is great and all but fully autonomous agents (AutoGPT et al.) aren’t happening any time soon. Compounding error is still too severe for long horizon, multi-step tasks. The agent “diverges” and gets stuck in a thought loop. Bullish on human-in-the-loop concepts though

2

0

9

@disiok

Simon Suo

2 months

wow elon delivers:

Tweet card media

GitHub - xai-org/grok-1: Grok open release

Grok open release. Contribute to xai-org/grok-1 development by creating an account on GitHub.

1

1

9

@disiok

Simon Suo

11 months

Checkout this super insightful talk about RAG from @lateinteraction

@jerryjliu0

Jerry Liu

11 months

The DSP project carries a lot of insights for improving RAG: 💡value of few-shot ex’s 💡declarative modules 💡compile an optimized system with distilled LM’s We had a GREAT time chatting about this and more w/ @lateinteraction on our latest webinar! 👉

0

11

71

0

5

9

@disiok

Simon Suo

5 months

2024 is gonna be the year of open source LLMs

2

0

8

@disiok

Simon Suo

11 months

Super excited for this! Come hang out with us at the @UCBerkeley AI/LLM hackathon

@chontang

Chon Tang

11 months

I'm super excited to announce - the best hackathon ever?! @UCBerkeley AI/LLM hackathon featuring @OpenAI : - $200k investment - 1k+ hackers Partners: @msft4startups @langchainAI @llama_index @mindsdb @pinecone @sequoia @lightspeedvp @mayfieldfund @FeatureformML What? Why?

Tweet media one

19

88

419

0

1

8

@disiok

Simon Suo

3 months

@jxnlco I hear something called if statements are blazing fast. Who are the investors?

2

0

8

@disiok

Simon Suo

11 months

@jxnlco @jerryjliu0 Just change your name to json instead of Jason

1

0

8

@disiok

Simon Suo

10 months

Disclaimer: This is an adhoc investigation, and I've only tried a couple of prompts. I also don't know whether replicate hosted model handles encoding differently. But the takeaway stands: watch out for subtle train/test mismatch bugs, especially when you don't own the API.

0

0

8

@disiok

Simon Suo

3 months

RAG = IR + NLP

Tweet media one

1

0

8

@disiok

Simon Suo

6 months

Season finale, back to where we started

@OpenAI

OpenAI

6 months

We have reached an agreement in principle for Sam Altman to return to OpenAI as CEO with a new initial board of Bret Taylor (Chair), Larry Summers, and Adam D'Angelo. We are collaborating to figure out the details. Thank you so much for your patience through this.

6K

13K

67K

0

0

8

@disiok

Simon Suo

2 months

Why is modal so cool

@modal_labs

Modal

2 months

Customer showcase: use Devin to write Modal code – it uses Modal, so you can use Modal while you’re using Modal! 🔄🤯

3

3

85

1

1

8

@disiok

Simon Suo

3 months

Do you ever wonder if you are training data quality constrained or brain size constrained? Just me?

1

1

7

@disiok

Simon Suo

18 days

@moinnadeem the beauty of speaking at macro / vision level is that it's not entirely falsifiable, so you can get away with anything

1

0

8

@disiok

Simon Suo

2 months

It’s over when the nerds start power lifting

@WuNeal

Neal Wu

2 months

Yes

Tweet media one

Tweet media two

270

345

8K

0

0

8

@disiok

Simon Suo

4 months

Startup is like padding a kayak on the ocean, with 10ft waves crashing, while aircraft carriers fighting each other nearby. Kinda fun though would recommend

0

0

8

@disiok

Simon Suo

5 months

You can sense the marketing pressure from the SoTA comparison

1

0

7

@disiok

Simon Suo

8 months

now this, is an exciting event with real signal. incredible lineup across model, infra, tooling, and application.

@robertnishihara

Robert Nishihara

@robertnishihara

8 months

Ray Summit this month will be 🔥🔥 🤯 ChatGPT creator @johnschulman2 🧙‍♀️ @bhorowitz on the AI landscape 🦹‍♂️ @hwchase17 on LangChain 🧑‍🚀 @jerryjliu0 on LlamaIndex 👨‍🎤 @zhuohan123 and @woosuk_k on vLLM 🧜 @zongheng_yang on SkyPilot 🧑‍🔧 @MetaAI on Llama-2 🧚‍♂️ @Adobe on Generative AI in…

8

45

208

0

2

7

@disiok

Simon Suo

6 months

Congrats on the launch! This looks incredible

@newcomputer

New Computer

6 months

Announcing our first product, Dot — an intelligent guide with incredible memory. Read the full story here:

209

507

5K

0

1

7

@disiok

Simon Suo

3 years

@sergioksas @RaquelUrtasun Paper & video:

Tweet card media

TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviors | Simon Suo

Simulation has the potential to massively scale evaluation of self-driving systems enabling rapid development as well as safe deployment. To close the gap between simulation and the real world, we...

www.cs.toronto.edu

0

4

7

@disiok

Simon Suo

4 months

In 2023 my Twitter feed shifted from academics to startup founders. Striking how everyone lives in their own bubbles & optimizes their own imperfect proxy metrics. Goodhart’s law Reminder that you get to pick your own metric.

0

1

6

@disiok

Simon Suo

4 months

Realized I’ve never fully appreciated the beauty of human language. It’s so perfectly imperfect. A single representation that allow us to express thought across various abstraction levels. Structured enough to be compositional building blocks, yet flexible enough for adhoc…

2

0

7

@disiok

Simon Suo

1 year

The history of ML feels like a pendulum swinging between domain-agnostic and domain-specific models. So far, domain-agnostic approaches have produced most big step changes, but what ends up getting deployed are highly optimized domain specific models.

@arankomatsuzaki

Aran Komatsuzaki

@arankomatsuzaki

1 year

Gorilla: Large Language Model Connected with Massive APIs Releases Gorilla, a finetuned LLaMA-based model that surpasses the performance of GPT-4 on writing API calls. proj: abs:

Tweet media one

18

192

747

0

0

6

@disiok

Simon Suo

11 months

Super useful project for comparing open source LLMs from ⁦ @anyscalecompute ⁩

Anyscale Endpoints

Anyscale Endpoints is a fast and scalable API to integrate OSS LLMs into your app. Use our growing list of high performance models or deploy your own.

www.anyscale.com

0

1

7

@disiok

Simon Suo

10 years

TIL the standard list of metasyntactic variables are: foo, bar, baz, qux, quux, corge, grault, garply, waldo, fred, plugh, xyzzy, thud

1

6

5

@disiok

Simon Suo

5 months

In a few years, kids grew up on post LLM internet are gonna start saying “as an OpenAI model…”

0

0

5

@disiok

Simon Suo

10 months

First, the "llama13b-v2-chat" refers to the fine-tuned variant of the Llama 2 model. While the base model simply completes text, the "chat" model is fine-tuned for conversations and instruction following. But how does it recognize messages and instructions?

1

1

6

@disiok

Simon Suo

3 months

Lemme check with llamaindex marketing department if we can buy a 0.05s Super Bowl ad next year. If anthropic can do it, why can’t we!

2

0

6

@disiok

Simon Suo

9 months

@OpenAI The launch blog () did a great job of highlighting use-cases and setting expectations. Fine-tuning helps with both quality/reliability and cost/latency. * quality/reliability: improve steerability, reliable output formatting, custom tone * cost/latency:…

GPT-3.5 Turbo fine-tuning and API updates

Developers can now bring their own data to customize GPT-3.5 Turbo for their use cases.

1

0

6

@disiok

Simon Suo

1 year

Tweet media one

1

0

5

@disiok

Simon Suo

8 years

Chilling at #12BARZ with the bae @cselina_ .

Tweet media one

1

1

2

@disiok

Simon Suo

5 months

Human engineer a tedious process Human engineer a model to reverse engineer the tedious process Why

@AndrewCurran_

Andrew Curran

5 months

Microsoft is training a custom, narrow-focus LLM specifically on the regulatory process for small nuclear plants. They need to build SMRs to power Bing's brain. MS expects the LLM to eliminate 90% of the costs and human hours involved.

Tweet media one

40

109

942

0

0

6

@disiok

Simon Suo

5 months

How do I get a 1-on-1 with @gdb and post a selfie on X

0

0

6

@disiok

Simon Suo

4 months

@jxnlco at one point in my life I thought happiness is about reading philosophy and grasping the truth of nature and self turns out it's just sleep, eat, and exercise well

1

0

6

@disiok

Simon Suo

18 days

starting to love the concept of agent-data interface draws analogy to HCI and encompasses RAG, text-to-sql, knowledge graph queries etc

0

0

5

@disiok

Simon Suo

8 months

Gonna tell myself to take a deep breath before doing anything important. Because if a LLM can do x% better, I can too!

3

0

5

@disiok

Simon Suo

10 months

So, does using the reference encoding make a difference? Let’s test it out. First: without any prompt structure… just doesn’t work at all.

Tweet media one

Tweet media two

3

0

5

@disiok

Simon Suo

9 months

@OpenAI There's also an updated fine-tuning guide with more details (). Some practical advices from the guide: 1. split fine-tuning dataset into train & test (fine-tuning job will provide stats on both) 2. check data formatting with provided script locally before…

Tweet card media

OpenAI Platform

Explore developer resources, tutorials, API docs, and dynamic examples to get the most out of OpenAI's platform.

platform.openai.com

0

0

4

@disiok

Simon Suo

6 months

2022 gives us FTX the movie 2023 gives us OpenAI the movie What a time to be alive

2

0

5

@disiok

Simon Suo

10 months

Second: using “Human: … Assistant: “ results in rambling, hallucinating multi-turn conversation

Tweet media one

Tweet media two

1

1

5

@disiok

Simon Suo

1 year

RAG gang

@nickfrosst

Nick Frosst

1 year

The term “hallucination” for large language models seems really misleading. All ungrounded generations are hallucinations, just sometimes you get lucky and what comes out of the model is true This is why retrieval augmented generation is so crucial for making this stuff useful

9

18

123

0

1

5

@disiok

Simon Suo

2 months

Cloud era: separation of compute and storage AI era: separation of reasoning and knowledge

0

0

5

@disiok

Simon Suo

3 months

My favourite model release of the year

Tweet card media

GOODY-2 | The world's most responsible AI model

Introducing a new AI model with next-gen ethical alignment. Chat now.

3

0

5

@disiok

Simon Suo

5 months

Undergrad, PhD, startup

@pmarca

Marc Andreessen 🇺🇸

5 months

2012-2016 was a distinct era. 2017-2022 was another. 2023- is new.

169

330

4K

0

0

5

@disiok

Simon Suo

9 months

PSA: @OpenAI fine-tuning is rate-limited to 1 active job per model per organization You can run 12 jobs per day though. 💸💸💸

Tweet media one

@disiok

Simon Suo

9 months

My notes on the new @OpenAI fine-tuning release: 1. currently limited to GPT-3.5 Turbo with 4K token (GPT-4, function-calling, and gpt-3.5-turbo-16k all coming this fall) 2. similar DX as old GPT-3 fine-tuning API, just list of chat messages instead of (prompt, completion) pairs…

Tweet media one

Tweet media two

5

6

34

0

0

5

@disiok

Simon Suo

4 months

crazy how "large" became a marketing term to the point where we now have "tiny L(arge)LMs" next stop: big TLLMs

0

0

5

@disiok

Simon Suo

11 months

Academic conference poster sessions vs. Startup conference screen sessions

Tweet media one

0

0

5

@disiok

Simon Suo

13 days

Pure behaviour cloning would’ve solved self-driving if not for the sophons

@ericjang11

Eric Jang

14 days

if your transformers struggle with NaNs after a certain parameter size, you may be under a sophon lock. Keep pushing, don't let them win!

8

10

209

0

0

5

@disiok

Simon Suo

3 months

Every entrepreneur out there reflecting if they are dreaming big enough

0

0

5

@disiok

Simon Suo

1 year

@karpathy What a legend. From ML recommendations () to LLM recommendations

Tweet media one

0

1

5

@disiok

Simon Suo

11 months

GPT Best Practices = just be a good PM!

Tweet card media

OpenAI Platform

Explore developer resources, tutorials, API docs, and dynamic examples to get the most out of OpenAI's platform.

platform.openai.com

0

0

5

@disiok

Simon Suo

5 months

In what universe would I want to use a projector instead of a screen

@SquawkCNBC

Squawk Box

5 months

FUTURE OF TECH: @Humane Co-Founders @bella_bongiorno and @imranchaudhri show off the AI pin –a new wearable device powered by artificial intelligence.

88

76

352

1

0

5

@disiok

Simon Suo

21 days

It’s gonna be weird when we start naming kids “Llama 3 David”

1

0

5

@disiok

Simon Suo

5 months

Water water water

@BasedBeffJezos

Beff – e/acc

@BasedBeffJezos

5 months

Hey everyone, serious post for a second. As you may have seen, I have unfortunately gotten my identity and stealth startup doxxed by reporters via voice forensics and web sleuthing. As the day has finally come, I thought I’d share more about who I am. I’ve kept my identity…

594

403

7K

0

1

4

@disiok

Simon Suo

4 months

@CalvinnChenn @jerryjliu0 Was convinced since we became friends in 2018

0

0

4

@disiok

Simon Suo

3 months

@EugeneVinitsky vehicle heading behaves like a usb stick. you only get it right on the third attempt.

1

0

4

@disiok

Simon Suo

18 days

@eugeneyan @BEBischof Would be fun to discuss how these archetypes fit in with the evolution of a early stage startup

1

0

4