
Agent{Force} Optimization R&D | Funnels Driven, Guardrail-First Deployment | Partnering with System Integrators, 💝 my playground... @3MoMLife @MyAiNativeTown
Bhiwani, India
Joined January 2010
- Tweets37.1K
- Following345
- Followers638
- Likes9.8K
Pinned Tweet
I think, I'll need a 100 agents foundation to manage by business operations.
I've got only 2. 98 more to go. 😳
i've got all the intelligence i need,
but i are probably stuck with some internal struggles...
Replying to @tantrumsofsadhu
Are you sure?
Let me start with this:
How many tasks a day take that 3-5 minutes?
You've got all the intelligence you need but you are probably stuck with some internal struggles.
3
An ideal conversation is one in which both parties enjoy, celebrate, and make the most of their interaction, while holding a strong belief in each other's efforts and intentions to enable, enrich, and bring out the best in one another's lives.
It is a space where curiosity is welcomed, differences are celebrated, disagreements are safe, and the possibilities created together exceed what either could have discovered alone.
definition of the term CONVERSATION:
Iteration #5
Every conversation is a highly scientific phenomenon, occurring simultaneously in multiple funnels and lives at the same time.
हर एक बातचीत एक बहुत ही साइंटिफिक घटना है, जो एक ही समय में कई फ़नलज़ और ज़िन्दगियों में होती है।
It involves multiple forces of nature, so that everyone involved can benefit as much as possible.
इसमें प्रकृति की कई ताकतें शामिल होती हैं, ताकि इसमें शामिल सभी लोगों को ज़्यादा से ज़्यादा फ़ायदा हो।
It depends on frequency, intention, and things known and unknown.
यह हमारी फ़्रीक्वेंसी, इरादों और न जाने कितनी चीज़ों पर निर्भर करता है।
Depth and mutual benefit come from the alignment of frequencies of everything you connect with, whether actively, passively, or as a second, third, or fourth-order effect.
और बात की गहराई और आपसी फ़ायदे हर उस चीज़ की फ़्रीक्वेंसी के अलाइनमेंट से आते हैं जिससे आप एक्टिवली, पैसिवली, या दूसरे, तीसरे, चौथे ऑर्डर के इफ़ेक्ट के तौर पर जुड़ते हैं।
7
Surviving public scrutiny at the scale of the country we have is such a marvel in itself.
You've got to have a thick skin & deep rooted principaled soul.
I wonder why @SarvamAI doesn't have a principaled agent like @boardyai which exists like talks and responds to the questions in balanced manners.
Representing an organisation is within AIs capabilities right?
Someone just needs to dare to build it like a true research project.
I am learning the basics of experimentation once again from an article @beh_zod wrote a long time ago.
None of my public experiments seem to have been designed for making the next decision, carefully... Yet. 😅
Hey, I've got skin in the game. My next iteration has maximum dots connecting all the past in various projects & disciplines & more proof of work. 🎯🤓
I wonder how much of the community building support @SarvamAI leadership can put behind my projects like @MyAiNativeTown to build a research oriented builders building culture over a long horizon.
Really don't understand why launching a token-factory product is getting so much criticism thrown @SarvamAI. I am obviously biased as an investor but I think people don't know just HOW MUCH some of the global players are making by hosting their global (incld chinese) models in India AND serving global customers from here, simply because they can't find GPUs elsewhere right now! You'd be shocked. And yes, goes without saying, the larger models are part of the roadmap, but just because a company launches product-x, doesn't mean that's where the launches are going to stop? There are so many parallel teams working on really ambitious products all launching at their own cadences. Genuinely confusing... All the other feedback around model latency, cost, performance etc are all really useful ones.
1
41
Hari ॐ Vashishtha 🤘 retweeted
Replying to @GulatiSinghJ
I've not been to most of them but i think my town has it's own unique character. 💝🎯🤓
I wonder how many in my town know it very well... Everyone knows it in parts.
There's no intelligence layer on the top of it, I wish to build one. 🎁
Hari ॐ Vashishtha 🤘 retweeted
0555am club anyone?
Be obsessed with “G”
Goals. Greatness. God. Growth. Gratitude. Grit. Good health.
Lessgooo 🚀
Hari ॐ Vashishtha 🤘 retweeted
Asking family and friends is how India works.
And our first agent for students in India is now WIP!
@GulatiSinghJ
#ComingSoon
'Main kisi ko jaanta hoon' (I know someone).
Everybody in India has that niece/neighbbour/family friend they know will get a job done that they don't understand.
Friction for us is never about intent but access and so adoption becomes a word in pitch decks and policy papers.
@runcompute, we just want everyone to have complete, free, trusted access. Your niece will not have to schedule you in anymore.
What opinions are you fighting today?
"How are you not outraged with <nationalism bullshit>"
"Psst.... I have a whatsapp group"
Every damn time.
And running inference with open models (regardless of country of origin) with data residency and billing is actually quite needed and a good gap to fill.
10
Asking the right questions...
Is part of the challenge.
Replying to @Sohail__shah_ @glnarayanan
You're right. I am curious though, how much private market funding has Sarvam raised and how much Government funding have they raised?
Do you have numbers? I would love to check something.
8
Maybe the eternal question i am pondering is:
How do i impact the culture of my town to be more empowering to each other?
Imagine a @boardyai like system for my town... Bhiwani, Haryana, India. 🇮🇳
Imagine: Everyone Knows Everyone.
Anyone can reach out to anyone who is willing to connect... {to be more concrete}
What a dream... @MyAiNativeTown 💝
AECI stands for Anthropic ECI (Anthropic’s version of the Epoch Capabilities Index).
It is a composite score that aggregates a model’s results across many individual benchmarks into a single overall capability number.
Anthropic created it as their internal fork of Epoch AI’s public Epoch Capabilities Index (ECI), which uses statistical modelling (similar to item-response theory) to stitch together dozens of tests of varying difficulty so progress can still be tracked after individual benchmarks saturate.
We can only improve what we can measure, right?
30
Hari ॐ Vashishtha 🤘 retweeted
Replying to @IntuitMachine
I'll pick up a few bits on those lines.
I've been looking for questions to be asked on similar lines.
Makes me think: Establishing the shared commitments that humans and machines need in order to act together.
Hey Carlos, your post helps me somehow but am not sure, where this is going so i am in a state to define the experiment, based upon @beh_zod's methodologies.
lnkd.in/p/gUsikYBa





















