Pratap Ranade
@PratapRanade
Followers
1,439
Following
487
Media
45
Statuses
330
Co-founder/CEO of @TheArenaAI . Prev CEO of @KimonoLabs (acq. @PalantirTech ) Triathlete + Physicist. @ycombinator @McKinsey , @Stanford alum.
New York, NY
Joined May 2012
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
Peanut
• 658923 Tweets
Georgia
• 462655 Tweets
Faker
• 243061 Tweets
Iowa
• 131917 Tweets
#StarAcademy
• 80552 Tweets
Penn State
• 75156 Tweets
Harrison Ford
• 72583 Tweets
Kelce
• 68473 Tweets
Selzer
• 55622 Tweets
文化の日
• 40754 Tweets
Fluoride
• 38493 Tweets
Vitória
• 33641 Tweets
#MostRequestedLive
• 26069 Tweets
Maluma
• 19360 Tweets
フォーエバーヤング
• 18591 Tweets
Nebraska
• 18350 Tweets
#全日本大学駅伝
• 16442 Tweets
Emerson
• 16222 Tweets
BCクラシック
• 15354 Tweets
#UFCEdmonton
• 15198 Tweets
Michigan
• 13440 Tweets
Big 12
• 13146 Tweets
Solari
• 12596 Tweets
ローシャムパーク
• 12500 Tweets
シャフリヤール
• 11755 Tweets
Carson Beck
• 11360 Tweets
South Carolina
• 11254 Tweets
Pinned Tweet
With
@ycombinator
demo day coming up this weekend, I wanted to reflect on what I learned from YC W14, 10 years on + another company in:
1. The 7 minute espresso rule.
Our first meeting with
@sama
lasted just 7 minutes. The batch hadn’t even officially started yet and my
31
57
533
1/ Huge thanks to
@garrytan
for sitting down with me to share how
@TheArenaAI
helps enterprises run like F-22 fighter jets by intelligently automating decisions like pricing
6
11
61
@sciencecohen
@imperialcollege
In case you're curious about how Imperial and the other teams are building their models, we just published a comparison of the approaches, and included open source code & data. Feedback welcome:
2
11
32
@adamscrabble
In case you're curious, here is a comparison of the different forecasts being used, and how each is sensitive to different assumptions. Interactive, so you can tweak yourself. Code & data is open sourced so others can extend: . Welcome your thoughts
0
10
29
I couldn’t be more excited to announce our collaboration with
@AMD
.
Over the past year
@TheArenaAI
has developed and deployed Atlas – the world’s first AI hardware engineer at AMD, a leader in high performance and adaptive computing.
The AMD team is using Atlas to enhance
1
6
32
I’m excited to share Arena’s $32 MM Series A led by
@Initialized
and Goldcrest, with participation from
@foundersfund
&
@flexport
. I’m so grateful to our investors for supporting us and to our incredible team for their dedication and tireless work!
2
4
22
@justin_hart
@BrendanEich
This may be interesting: We published a comparison of 3 modeling approaches (including Murray's ). You can tweak the assumptions yourself to see how things change. Code & data open sourced:
Feedback welcome!
0
10
18
I’ve been thinking a lot about
@parkerconrad
’s compound startup idea applied to AI.
Many are still trying to apply the old SaaS playbook to AI, but that’s the way of the past. The way to build a generational company in AI is more of a “joint-mission” model.
Big enterprises hold
0
2
19
When is the peak? We just published a comparison of 3 different COVID-19 forecasting models. We open sourced all the code & training data - it updates daily. Hope some of you find this useful and would welcome feedback on how to make it better:
0
9
18
Here’s the full version of our pitch if it’s helpful to the current batch / anyone else:
2
0
17
1/ it’s launch day at
@TheArenaAI
! Inviting you to try to beat our AI in a sim we’re unveiling today called Retail Tycoon 🛒📈
2
1
16
Part 3 of my interview
@jevering
, head of eng
@instagram
for
@enigma_data
on finding stellar talent:
1
6
15
@Initialized
@foundersfund
@garrytan
@AldaLeuDennis
@LM_Braswell
@zoebeutel
@mwseibel
@taps
Ultimately, I think we all do startups not because they are easy, but because they are worthwhile.
Our neon shoelaces remind us of the privilege we have to be racing in the Arena versus watching from the sidelines.
0
4
11
On their first day, every new teammate who joins
@TheArenaAI
receives a pair of these bright neon laces.
Curious to know why?
1
1
10
When we raised our Series A, we wanted to remind ourselves that raising funds itself is not a sign of success, but rather an invitation to the race.
So to celebrate, all of us signed a pair of neon racing spikes.
1
0
10
🚨PSA for the NYC AI community🚨
So many of you asked for this, so we’re opening up our bi-weekly ML journal club to anyone who’s interested in NYC. Thursday, March 14th, at our Arena office. We'll be diving into the guts of Gemini. And there will be pizza. Space is limited –
1
1
10
Sat down with
@jevering
, head of eng at
@instagram
to talk about quantum management for
@enigma_data
. Notes here:
0
5
9
@typesfast
- inspired by your and
@flexport
’s speed at brining PPE to our medical workers. Lucky to be able to donate — friends, please join me, and help!
0
2
8
In the AI world, I’m seeing a failure mode where people think that all data is useful information. We chase mountains of data – presuming the bigger the mountain, the higher the value, and ultimately sit on a pile of noise as a result.
2
1
8
@enigma_data
is searching for a talented Director to lead our data acquisition efforts and launch our first expansion office outside of NYC! Apply now if you're interested or refer someone who is passionate about data! Open to all cities.
0
3
7
We’re building a probabilistic knowledge graph of the real world at Enigma. Excited to add fuel to the fire with our latest 95M round. Come build it with us!
We’re changing the way people interact with data. Today, we’re excited to announce we raised $95M, which will go a long way to expand our core offerings, bring on more top-tier talent and further invest in product R&D.
0
10
34
0
0
7
Here's part 2 of my interview series with
@jevering
, head of eng
@instagram
for
@enigma_data
. This time we discuss wardialing.
0
2
5
For machine learning folks in NYC looking to do something exciting and early stage with cutting edge reinforcement learning. We're hiring!
0
1
5
Silicon Valley is unique. Every country is trying to build their own and nobody has come close. It’s the nursery that gave birth to Google, Microsoft, Apple & SpaceX. Letting SVB die will be a massive long term blow to American innovation. Letting SVB die is not an option.
#SVB
1
1
6
“Geniuses Are Lazy — and You Should Be Too” - my thoughts on how physicists turn hard problems into easy ones, and how everyone can do the same.
0
4
6
The more AI folks in NY, the better
Move to NYC! We have bagels and culture and public transportation and
@srush_nlp
and bagels!
7
8
138
1
1
6
One of my favorite Star Trekisms: the distinguishing quality of humans is that they aspire.
These days it feels like a lot of people try to prove they are better than others, but few aspire to simply become better.
0
0
5
@nathanbenaich
@stateofaireport
I like this - A100 GPUs are a good proxy. Two things this might miss: 1) the bigger shops running on TPUs / other non GPU architectures and 2) developments in low latency edge AI
1
0
5
Multimodal AI will be a step-change in how and where AI can be applied. Integrating modes beyond text and images will open up entirely new applications in the physical world. What is multimodal AI, and what will these opportunities look like?
Read more of my thoughts here:
0
2
5
Excited to share some of the entity resolution work we've been doing
@enigma_data
by
@superlativest
!
0
3
5
Remember the scene from the Matrix where Morpheus fights Neo in a virtual dojo? Keanu Reeves as Neo opens his eyes and says that iconic line “I know kung fu”.
They fought in simulation, but the skills they acquired transferred to reality. Could we build simulators to teach
1
0
3
I’m on the plane home from 2 weeks embedded with customers across the world. It was a powerful reminder that:
1) Ambition isn’t exclusive to Silicon Valley
2) Enterprise CEOs want to embark on “Apollo-style” missions but need deep technical partners to take big swings, not LLM
1
2
4
6 Lessons I learned from
@PalantirTech
:
1. Access to problems is earned
2. STFU and Take Notes
3. Learn with a thief’s mindset
4. Get your a** on a plane
5. Look at raw data
6. Build the whole car
0
4
3
Sad, sad day. Without the H1B program, I wouldn't be here. I wouldn't have a chance to chase the dream.
Nor would we have many great innovators – from
@AndrewYNg
to
@elonmusk
.
The suspension of the H1B visa program is bad for the US, bad for innovation, and will shatter dreams and disrupt lives. As a former H1B visa holder, my heart goes out to all the families affected.
228
3K
18K
0
1
4
Building in AI I is more than a step change.
It’s a phase change.
It’s like we’re moving from a solid to a liquid. New laws of physics apply.
Building in AI requires us to think differently about strategy.
Borrowing a concept from physics – the light cone – can help:
2
0
4
This man has inspired me more than perhaps anyone else. Excited to continue learning from you,
@SirPatStew
It is an unexpected but delightful surprise to find myself excited and invigorated to be returning to Jean-Luc Picard and to explore new dimensions within him. Read my full statement in the photo.
#StarTrek
@cbsallaccess
Photo:
@shervinfoto
5K
34K
133K
0
0
4
Enterprise SaaS thinking isn’t going to work for AI.
The enterprise SaaS playbook that was written over the last 20 years is narrow and focused on incremental efficiency gains delivered through scalable point solutions. It is limiting and zero-sum.
AI is positive-sum. Its full
0
0
4
Thanks
@kinlane
- always fun catching up! Opportunities in API Deployment For Non-Developers -
http://t.co/EQT2sdF6TU
via
@apievangelist
1
1
4
@LM_Braswell
@TheArenaAI
Thanks for being one of our earliest supporters Leigh Marie - glad you’re part of the Arena team!
1
0
3
2/ Garry has been someone I have looked up to for years - he was one of the first partners I met at
@ycombinator
with my first company
@kimonolabs
and he’s been an investor and champion for both Kimono and Arena
1
0
3
4/ it’s hard to get a feel for abstract AI, so we built a demo LXM using public data –– it helps you choose where to place a store. Pick a spot and the LXM will simulate the full year of credit card spending at that location and other stores in that city
2
2
3
This.
We need a lot more AI sensors - and not passive sensors. Active sensors that actively maximize information capture from the environment
E.g. our eyes aren’t passive sensors. They are active. If your peripheral vision perceives movement, your eyes swivel to look at it and
"A 4-year-old child has seen 50x more information than the biggest LLMs that we have." -
@ylecun
20mb per second through the optical nerve for 16k wake hours 🤯
LLMs may have consumed all available text, but when it comes to other sensory inputs...they haven't even started.
54
290
1K
0
2
3
If you want to virtually join our ML journal club - here are two papers that we spent time on:
1. The new "og" Mamba:
2. Vision Mamba:
Will State Space models replace transformers?
📸 from this week’s ML journal club
@TheArenaAI
where our ML scientists had a heated debate on the topic.
It’s so gratifying to be working in applied AI right now, with a team of incredibly talented people who embrace the beginner’s
0
0
1
0
0
3
3/ I love the analogies he uses here to describe our tech - AI pricing has the potential to be gamechanging for our customers - to the tune of billions of additional revenue for some of the biggest businesses in the world…
1
0
3
Great piece by
@scottehartley
: What Cyborgs and Moneyball Have in Common
@kimonolabs
@inc
http://t.co/C1OPNwOXO4
via
@Inc
0
2
3
@nikillinit
@kimmaicutler
@ycombinator
@sama
@paulg
@gralston
@collision
@garrytan
@mwseibel
We were both into no-code before no-code was a thing!
0
0
3
0
0
2
1
0
2
@johnalxndr
@ycombinator
@sama
@paulg
@gralston
@collision
@garrytan
@mwseibel
Means a ton. Even 9 years later
0
0
2
0
0
1
@FiveThirtyEight
Couldn't agree more! We just wrote up a comparison of 3 different COVID-19 forecasting models being used with open source code & training data. Would love your thoughts: ,
@waiting4spark
0
0
2
1. Access to problems is earned. Learned this from CTO
@ssankar
. When working with enterprises that are large/global/complex, you can’t understand deep problems outside in & you won’t get invited to solve it at it. You have to earn the right. True with customers & inside the co
1
0
2
One of our guiding principles at
@TheArenaAI
is Real World > Lab. AI has to step into the real world and actively acquire new information to move us toward true AGI. Great example of this from
@keerthanpg
and
@svlevine
You cannot solve self driving by driving in a parking lot and you cannot solve robotics in a lab.
Much before we build physical AGI, we gotta force robots to leave the lab for the real world and collect lots of data about the nature of physics, reality and humans.
16
37
401
0
0
2
@sama
a founder’s founder.
Sam has been sending stuck startups money today with no docs, just saying “send me back whatever you can whenever you can”. What a legend 🥲
161
361
6K
0
0
2
0
0
2
Love your
@FiveThirtyEight
Markov bots
@WaltHickey
-
http://t.co/IkzsITbnxH!
Stoked that
@kimonolabs
was able to help bring them to life
0
0
1
2/ Like everyone, I couldn’t be more excited about the (well-deserved) hype for generative AI and Large Language Models. At
@TheArenaAI
, we work with Fortune 500 businesses – the data we harness isn’t a sequence of words or pixels, but a sequence of SKUs and credit card swipes
1
0
2
1
0
2
In today’s substack, thought I’d unpack a few of the reasons that transformers are quietly revolutionizing AI, and their similarities to our own 🧠
0
0
2
@AravSrinivas
@tobi
@naval
@rauchg
@balajis
Congrats
@AravSrinivas
. Inspiring to watch the pace at which you and the
@perplexity_ai
team have been shipping
0
0
2
Great article by
@paul_arnold
on how to grow your startup 500% in months:
http://t.co/kaIQ9kmMgk
0
0
2
1
0
2
Light cone strategies ask for a vector.
Instead of a point, draw a radial line pointing out to the application boundary to put yourself on the glide path of the cone.
If you can grow along the cone, your circle of opportunity expands + pushes outward with more AI innovation:
2
0
0
We can do better. We need to rethink our sensors, and also how we price data. Entropic pricing and entropy-maximizing sensors could change the game for AI – enabling it to leap out from text and images on the internet and into the real world.
1
0
2
5/ you can read more about the tech behind Retail Tycoon here:
1
0
2
3 ways to measure your adaptability — and how to improve it (Natalie Fratto | TED Residency) via
@TEDTalks
0
0
2
@nathanbenaich
@stateofaireport
That definitely tracks with what we are seeing. Thanks for putting together and sharing this data — quite a thoughtful analysis
1
0
2
RT:
@BeyondAcademia
my thoughts on the journey from a PhD to a startup-
http://t.co/ss3whdDXhg
0
1
1
3/ The properties that make large transformers ideal for language also apply to many sequential decisions that are fundamental to how businesses are run – we call these foundation models LXMs
1
0
2
@COVID19Tracking
Thank for collecting all this data! Super useful. We used it to create a comparison of 3 different forecasting methods. Included open source code & other training data. All updates daily. Would love feedback if it is useful/interesting.
0
0
1
@kaggle
Love this! Our team of data scientists just published a comparison of 3 different COVID-19 forecasting models. We open sourced all the code & training data - it updates daily: - would love feedback from the DS community
0
0
1
0
0
1
Awesome initiative from
@typesfast
and
@flexport
. Our entire team at is donating to help - join us!
0
0
1
@paulg
PG - we just published a comparison of COVID-19 forecasting models, including IHME (UW)'s. Charts are interactive so you can tweak the assumptions. Open sourced all the code & training data. Updates daily:
Would welcome any feedback
0
0
1
@michelletandler
@kimonolabs
check out
http://t.co/zEdu52V1cO
#builtwithkimono
and
http://t.co/dZcECTydK7
for ideas!
1
1
1
The difference between a podium finish & an average race comes down to a series of tiny choices to push past pain.
The decisions that matter most are the ones you make when it hurts most. With lactic acid burning in your muscles and every fiber telling your body to quit.
0
0
1
@MichaelLFaye
@GiveDirectly
@PaulFNiehaus
inspired by your quick response to use GD to distribute money to families hardest hit by COVID. Lucky to be able to donate — friends, join me and help:
0
0
1
1
0
1
@ssankar
@SchimpfBrian
6. Build the whole car. Build a really crappy v0 of the entire thing end-to-end on day 1. Then iterate on the components vs the other way around. Building the whole car first results in a better product, shipped faster
1
0
0
The tradition comes from triathlon, which I raced competitively in college and graduate school.
In tri, transitions are everything. You have to be quick to rip off your wetsuit and get on your bike, and fast to swap shoes and switch from the bike to the run
1
0
1
Will State Space models replace transformers?
📸 from this week’s ML journal club
@TheArenaAI
where our ML scientists had a heated debate on the topic.
It’s so gratifying to be working in applied AI right now, with a team of incredibly talented people who embrace the beginner’s
0
0
1
@arpitrage
Great work from the IHME! We just released a comparison of 3 different forecasting models (including the IHME approach), with interactive charts so you can change the assumptions yourself. We included open source code & data. Would love feeeback!
0
0
1
To adapt the light cone for AI strategy, let’s replace the speed of light with the speed of AI research.
The slope is steep and the cone is wide because AI research is moving quickly.
The cross-sectional circle that it maps out on the present is the application boundary:
1
0
1
It’s time to build!
0
0
1