currently glowup health + @livetok_ai | former - ceo @hopin, founder @hopinsession, @tokbox, @microsoft, @stanford
"remote"
Joined December 2008
- Tweets1.6K
- Following300
- Followers1.2K
- Likes872
Stop the guesswork loop with Claude Code! 🔥
The AskUserQuestion tool lets Claude pause mid-task, hit you with smart multiple-choice questions (context-aware options + descriptions), and only then write code that actually matches what you want.
Game-changer for spec-based dev leading to fewer rewrites.
Good read here atcyrus.com/stories/claude-c…
#ClaudeCode #AICoding #DevTools
If this really is the future @elonmusk describes, the AI + robots → abundance really unfolds, then why is government efficiency (DOGE) such a high-leverage problem?
Isn't it almost a non-issue if this new reality & boom in productivity unfolds?
AI Voice Agents with one line of code
producthunt.com/products/liv…
@anarchyco and i have been working on a side quest
an AI that answers customer questions on WhatsApp — then accidentally made something that lets you add AI chat to any
website with one line of code
here's what happened 🧵
@anarchyco and i have been working on a side quest
an AI that answers customer questions on WhatsApp — then accidentally made something that lets you add AI chat to any
website with one line of code
here's what happened 🧵
the actually cool part:
same AI works across everything
- website
- whatsapp
- sms
- phone calls
customer starts chatting on your site, continues on whatsapp later. AI remembers everything.
anyway if you want to try it:
livetok.ai/widget-landing
or just paste that script tag anywhere and see what happens
lmk what breaks
Badri Rajasekar ✨ retweeted
I got this popup with AI-generated backgrounds in Google Meet today...
I don't know if it is new or it has been there for a while but reminded me when I built that feature a Sunday morning at Starbucks and we released it the week after.
In @HopinSession you can customize tons of things including many images (backgrounds, logos, virtual background...). Just hacked a quick demo to get infinite AI generated options, wdyt?
Next is to use the conversation, mood, weather... to autogenerate them :)
Greatest band that ever existed - if you disagree you are musically stunted ! Vulfpeck
youtu.be/C430AWTvMPs?si=pqvW…
Badri Rajasekar ✨ retweeted
Even in Silicon Valley, most people think that AI Apps/Agents are built just on top of leading models like Claude or ChatGPT.
The truth is that under the hood, more than 50% of LLM calls are actually made to cheap+open LLMs.
This effectively means that Chinese models like Kimi, DeepSeek etc. are powering most AI apps. There is no US alternative today. This should terrify you.
We need models that are built in the USA for the cheap+open use case.
This is why DeepCogito's announcement yesterday is so important.
Let us cook.
xitter.cf/drishanarora/status/19…
Today, we are releasing the best open-weight LLM by a US company: Cogito v2.1 671B.
On most industry benchmarks and our internal evals, the model performs competitively with frontier closed and open models, while being ahead of any US open model (such as the best versions of OpenAI’s GPT-OSS, Nvidia’s Nemotron and Meta’s Llama).
We also built an interface where you can try the model (it’s free and we don’t store any chats): chat.deepcogito.com
Additionally, you can download the model on @huggingface, or try it out on @openrouter, @togethercompute, @FireworksAI_HQ , @ollama cloud, @runpod, @baseten, or run it locally using @ollama or @UnslothAI.
This model uses significantly fewer tokens amongst any similar capability models, because it has better reasoning capabilities. You will also notice improvements across instruction following, coding, longer queries, multi-turn and creativity.
📌 Model Weights: huggingface.co/collections/d…
📌Openrouter: openrouter.ai/deepcogito/cog…
📌 HF Blog: huggingface.co/blog/deepcogi…
Some notes on our approach + design choices below 👇
Surveys are broken because they stop at what — not why. We’ve been forcing humans to talk to spreadsheets for decades. 🚨
The next generation of research tools won’t look like forms. They’ll feel like conversations — adaptive, curious, emotionally intelligent.
When an AI can listen, probe, and ask the right follow-up (“why did you feel that way?”), response rates triple — and the data becomes storytelling, not statistics.
And when those conversations happen where customers already are — WhatsApp, SMS, voice — the friction disappears.
A voice agent that talks like a person and listens like a researcher turns feedback into dialogue, and dialogue into trust.
The future of insights isn’t about collecting more data — it’s about earning more honesty.
Check out livetok.ai/cases/surveys
Badri Rajasekar ✨ retweeted
I call almost everyone.
Exponentially more efficient and can communicate on walks, in Ubers, etc.
Yesterday, someone I hadn’t responded to yet shot me a call out the blue and it saved us 30+ min of scheduling and doing a real “meeting”. So good.
"ALL net wealth since 1926 created by 3.44% of companies..." - feels like a classic argument both for and against indexing
Indexing works because you own the winners.
Indexing sucks because you also own 500 companies doing cardio for nothing.
This is wild:
ALL net wealth in the US stock market since 1926 has been generated by just 3.44% of companies.
To put this differently, ~97% of all stocks have barely contributed to long-term shareholder wealth creation.
The top 1.88% of companies reflect 90% of total gains.
Interestingly, just 0.26% of firms have created HALF of all wealth.
This highlights the extreme concentration of stock market returns in top-performing companies.
Market wealth is heavily skewed toward a very small minority of companies.
Badri Rajasekar ✨ retweeted
It’s been inevitable for a while now, but OpenAI is going to build an OS.
Also notable, much of their team, especially in design, worked on similar OS-like efforts at Facebook, like Facebook Home.
We’ve acquired Software Applications Incorporated, the maker of Sky—a natural language interface for Mac.
openai.com/index/openai-acqu…
Badri Rajasekar ✨ retweeted
Every time a new AI interface comes out, I try it. I'm convinced that the right conversational interface for AI will have tremendous demand and will change the way we interact with the digital world.
