Nikita Namjoshi
nikitanamjoshi.bsky.social
Nikita Namjoshi
@nikitanamjoshi.bsky.social
very human
explaining ai @ google deepmind | co-host theaustinsalon.com
nikitanamjoshi.substack.com
update: excited to be joining Google DeepMind 🪄 to help devs build with gemini + keep you up to date on all the most interesting ai research.
April 15, 2025 at 5:44 PM
two years of The Austin Salon ✨ with @dalequark.bsky.social
10 evenings spent discussing psychology, neuroscience, food equity, primatology, friendship, ai, and most of all community 🤍
April 9, 2025 at 4:16 PM
I see you Gemini 2.5 Pro Experimental 👀
aistudio.google.com/app/prompts/...
Sign in - Google Accounts
aistudio.google.com
March 25, 2025 at 5:40 PM
Who is going to be at the Build with AI Austin event on Thursday? 🙋🏻‍♀️ Got to try out the codelab this morning and can confirm it is very exciting.

rsvp.withgoogle.com/events/build...
Build with AI
rsvp.withgoogle.com
March 18, 2025 at 7:15 PM
🩶🐦‍⬛ sample the voices here: cloud.google.com/text-to-spee...
March 17, 2025 at 4:01 PM
everyone is talking about scaling laws, test time compute, and thinking models, so I wrote something for the uninitiated.

www.linkedin.com/pulse/thinki...
March 13, 2025 at 4:20 PM
open models are the best models 😊
Gemma 3 is here! My Gemma 3 highlights:

- 🚀 Gemma 3 delivers state-of-the-art performance for its size.
- 🇳🇱 Pre-trained support for over 140 languages.
- 🪟 128k-token context window.
- 🩷 Comes in three sizes: 1B, 4B, 12B and 27B.

Read more in our launch blog below.

#Gemma3 #Gemma #Google
March 13, 2025 at 1:21 PM
LLMs evaluating LLMs 🪄 a quick read for anyone who wants to understand the "autorater"
www.linkedin.com/pulse/llms-e...
March 11, 2025 at 3:06 PM
very compelled by this take. "until we find a way to incentivize them to question their knowledge and propose ideas that potentially go against past training data, they won't give us scientific revolutions yet."
I shared a controversial take the other day at an event and I decided to write it down in a longer format: I’m afraid AI won't give us a "compressed 21st century"

Here: thomwolf.io/blog/scienti...

It's an extension of this interview discussion from the AI summit: youtu.be/AxBd3G0lFLs?...
March 11, 2025 at 2:57 PM
Reposted by Nikita Namjoshi
If you’re writing a book right now you just have to pretend you’re a surgeon. Can’t look at your phone during surgery.
March 10, 2025 at 9:27 PM
preparing my arguments now.... 🔥
March 10, 2025 at 11:36 PM
🌬️ My thoughts on this week's executive order on AI + why AI alone won't solve climate change.
nikitanamjoshi.substack.com/p/why-does-m...
Why does my AI have a carbon footprint?
Just because it's in the metaverse doesn't mean it's not contributing to climate change 🫠
nikitanamjoshi.substack.com
January 16, 2025 at 10:23 PM
Codestral 25.01 from Mistral AI is now available on Google Cloud as a managed endpoint. 🪄 No need to provision or manage infrastructure.
January 13, 2025 at 11:41 PM
I really wish the part about a "murder victim’s corpse would spontaneously spout blood in the presence of its killer." was real 🥲
To early scientists, “the world teemed with the uncanny: witches, unicorns, mermaids, stars that foretold the future, base metals that could be coaxed into gold or distilled into elixirs of eternal life.”

🦄 Read my latest for Nautilus:

nautil.us/how-the-occu...
How the Occult Gave Birth to Science
For scientists of yore anything—from mermaids to alchemy—was on the table.
nautil.us
January 13, 2025 at 9:52 PM
ai face is the new instagram face
nikitanamjoshi.substack.com/p/the-age-of...
The age of AI Face
AI generated influencers and our obsession with perfecting the female body
nikitanamjoshi.substack.com
January 13, 2025 at 5:26 PM