Enabling {VCs/PEs/Family Offices} manage deal flow, portfolio, fund and LP's through Taghash (https://xitter.cf/t.co/ddBxFv1YFt) software. Founder/Cook @taghashhq
Mumbai/Udupi/Sitamarhi
Joined February 2008
- Tweets122K
- Following2.1K
- Followers6.3K
- Likes35.5K
Pinned Tweet
Big news from our side!
On this auspicious occasion of Dussera, we are thrilled to launch Brain, an AI agent built directly into Taghash! 🚀
After years of helping investment funds consolidate their deals, portfolios, investor records, and fund operations, we are taking things to a whole new level. Now you can put all that valuable institutional knowledge to work through a simple, natural conversation.
We compiled a short video to share Brain's capabilities :)
We're rolling it out to small cohort of Taghash customers this week, if you'd like to be one of those please reach out to us.
Thanks to the entire team at Taghash for pulling this off. And finally, this wouldn't have been possible without our customers and their desire to progress in the agentic era.
This is just the beginning. Even more exciting stuff on the way!
Atul Jha retweeted
Any job performed on the computer has this risk now: one smart person using AI can perform the work of 10 people.
The risk is getting ever more real.
This does not mean 90% of people working on a computer will have no work at all.
Industries that previously could not afford to hire software engineers or digitial marketers could afford them now, for example.
What it does mean is a rebalancing across the economy and that is painful when we go through it.
The only thing we can do is to accept that it is coming and be prepared for it.
Atul Jha retweeted
I have had something like this running for a few months now.
Striped back to basics: github.com/kbsingh/llm-servi… comments welcome
We just need a consistent spec.
The more I read news the more I feel we are in retarded era in all aspect. From globalisation to pollution to geopolitical tension to loss of job because of AI.
Atul Jha retweeted
So they pay you a $150K salary just to be a middleman between Claude and your boss?
Atul Jha retweeted
We at @suvanventures recently researched 25 AI-native companies to understand what enables them to scale beyond $5M ARR with relatively lean teams. Based on this research, we developed the Lean AI Scaling Framework, identifying seven key drivers of growth. Thank you @YourStoryCo for featuring our Framework. The link is below:
yourstory.com/2026/10/lean-a…
🤖 Made with AI
Atul Jha retweeted
Fuck you, @ArtificialAnlys
Your new index says Claude Haiku 5.5 is smarter than DeepSeek V4.1 Flash 1017 and GLM-5.3.
Haiku is Anthropic's small cheap model. DeepSeek 1017 is the model Cursor put on its cost frontier this month, and Cursor says no model scores higher at those prices. I run it every day on my 4x 3090. I dont believe for a second that Haiku beats it.
Look at how they got there. 60% of the index now runs on private tests nobody outside can see. GDPval-AA, SciCode and the long context test are gone. In comes a private coding benchmark from Cognition. And the open weights gap goes from 3 points to 12 in 1 release, 4x. Released open models are now 26 points behind the top.
1 day before v5, on the old version, they gave Ling 3.1 Flash 1019 a 44. Now the best released open model gets 32.
The weights didnt change. The tests did, and you cant see the tests.
This is a bought benchmark. A leaderboard where you cant check a single question, that widens the gap 4x overnight and puts Haiku above DeepSeek, is marketing for closed labs. Its not measurement. I hope at least you were paid enough money to publish such garbage. Liars.
Publish the questions or stop calling it an index.
Announcing Artificial Analysis Intelligence Index v5, featuring 2 new benchmarks and 3 upgrades. For the first time, the majority of the Index is built on private datasets
The Artificial Analysis Intelligence Index is our composite measure of model capabilities across tasks reflecting real-world usage today. v5 is our biggest upgrade since v4 launched in January. With v5, 60% of the Index is now based on evaluations with private test sets or held-out answers.
Alongside Intelligence Index v5, we’re launching the Artificial Analysis Progress Index v1, which tracks model capability over time on one continuous scale.
New to the Intelligence Index:
➤ FrontierCode 2.0: Cognition's newly-launched private benchmark of real-world coding tasks, graded on whether a repository maintainer would merge the agent's change
➤ Terminal-Bench-Science: 70 agentic scientific research tasks, run in a terminal environment
Upgraded:
➤ HLE-Diamond (replacing Humanity's Last Exam): a refined 1,000-question version of Humanity's Last Exam, split evenly between reasoning and knowledge questions
➤ AA-Omniscience v1.1: An updated dataset with saturated and public set questions removed, new questions added, and ambiguous or erroneous questions revised
➤ CritPt v2: a private, 40-question hard subset of CritPt v2, revised by the CritPt authors and run with a code-execution tool
Staying in the Index: AA-Briefcase, AutomationBench-AA, GDP.pdf and Terminal-Bench 4.0
Removed from the Index: GDPval-AA v2.1, SciCode, and AA-LCR v1.1
Key results:
➤ Claude Opus 5.5 (max) leads at 58, ahead of GPT-6 Astra (max, 57) and GPT-6.1 Sol (max, 55): Anthropic and OpenAI hold the top five models. Gemini 4 Argon (high) is next at 50
➤ The open weights gap widens from 3 points to 12: Kimi K3.1 Pro (weights not yet released) is the top open weights model at 46. Among released weights, DeepSeek V4.1 Flash 1017 (max) and GLM-5.3 (max) lead at 32 - 26 points behind the leader and below Claude Haiku 5.5 (max, 36)
➤ Cost per task varies by more than 13x among the leading models: of the 10 highest-scoring models shown, GPT-6.1 Terra (max) is the cheapest at $0.96 per Intelligence Index task and Claude Fable 5.1 (max) the most expensive at $13.04
Atul Jha retweeted
JUST IN: 🇷🇺 President Putin says Russia is proud to be developing its own competitor to Elon Musk's Starlink satellite internet.
Atul Jha retweeted
Was trying my hands on @Emergentlabs lately and made this tram ride in San Francisco
Always loved that thing in SF.
Pretty cool output I would say.
sfride-emergent.com
Atul Jha retweeted
It took me a while to realize that in Silicon Valley, “He’s good at fundraising” is often meant as an insult 😅
Atul Jha retweeted
i opened up an AI's brain tonight with @_lyraaaa_'s neuron explorer
qwen 3.5 4b has 294,912 neurons
push just one and it becomes someone else
here's who i found 🧵
In China, the AI bubble has already burst << US has not even started.
youtube.com/watch?v=uU2WeAX2…
They asked us for harness, we made harness defined agentic workflow. One system, no band-aid and dozen of plugins.
