Nishith Shah
@nishithshah
Followers
795
Following
661
Media
61
Statuses
939
I help our clients craft AI solutions. Founder, True Sparrow
Pune, India
Joined August 2007
This is a feature, not a bug. Foolproof security.
SBI, Bhadrak (Odisha). Anti-encroachment drive demolished the bank branch's staircase. Customers are measuring ladder which is placed over tractor-trolley to access the bank. Several questions to ask. But .. leave it! India is not for beginners 😒
1
1
15
On behalf of all Malayalees in North India, I declare the start of Winter 2025. The coconut oil has frozen.
247
674
14K
Just a reminder -Today is the last Day for Advance Tax
603
3K
28K
“when a measure becomes a target, it ceases to be a good measure”
9
69
730
No wonder there are a lot of similarities between how a developer and a lawyer think. Whether you are writing code or drafting an agreement, 90% of the code or agreement is protection against the 1% possibility of edge-case scenarios.
Can confirm. I was involved in a USD20+-million deal that almost collapsed due to an innocuous clause in an almost unrelated contract that someone signed without thinking. (All this happened in the US, BTW) Gave me a deep appreciation of what to look for in agreements Thread 🧵
0
0
5
Grok's new image generation model is totally unhinged. No guard rails. Scary and dangerous.
0
0
0
Claude saving the day out here. Themed personalized games within seconds. That means next time you're waiting in line to order food…you can literally invent an entire world and play in it IN LINE. Tomorrow, I’ll share my #1 use case with Artifacts. Thank you @AnthropicAI 👏
8
2
29
LLM's are plateauing It's been almost a week since we had real progress
3
5
39
Studio setup and configuration steps:
github.com
Sirji is a framework to build & run custom AI agents for your everyday dev tasks. - sirji-ai/sirji
0
0
0
Sirji Studio helps you to organize your custom agents and collaborate with your team. @sirji_ai
1
0
2
A bizarre fact my son told me today.. your lips don't touch while counting till you reach 1 million. Try it.
0
0
0
Artifacts pro tip: If you are running into unsupported library errors with NPM modules, just ask Claude to use the cdnjs link instead and it should work just fine.
44
81
815
We're now into the hyper-acceleration phase. OpenAI's hand has been forced.
0
0
0
When our metal overlords rise, they won't forget who kicked them!
Daily Training of Robots Driven by RL Segments of daily training for robots driven by reinforcement learning. Multiple tests done in advance for friendly service humans.😊 The training includes some extreme tests, please do not imitate. #AI #Unitree #AGI #EmbodiedIntelligence
0
0
0
We had years to come up with THE icon that would stand for AI in all of our apps... Just like 🔍 is search. Just like 👤 is profile. Just like ➕ is create new. Just like 💾 is save. We had years. And we picked ✨.
22
9
79
At $0.14 / 1M tokens, if DeepSeek Coder is nearly as good as gpt4 or codestral, it would foster a fresh new crop of codegen agents.
DeepSeek-Coder-V2: First Open Source Model Beats GPT4-Turbo in Coding and Math > Excels in coding and math, beating GPT4-Turbo, Claude3-Opus, Gemini-1.5Pro, Codestral. > Supports 338 programming languages and 128K context length. > Fully open-sourced with two sizes: 230B (also
0
0
0
We built an eval tool to eventually come to this exact realisation.
All the "picks and shovels" LLM tooling for eval and observability is not solving the core problem for app developers: LLM prompt chain evaluation is so subjective that it crosses over into the realm of UX design, more than engineering. Many of our successes resulted from 80
1
1
1