Shrey

Optimist Product Manager, figuring out life by exploring Philosophy, watching Cricket and reading ML blogs.

Product @Google 💼 🇺🇸, 🏠 🇮🇳.

[Shrey @shrey_g]

@GergelyOrosz @jon_roelofs Try "Ask Maps" feature on Google Maps. Its pretty good for these sort of things.

[Shrey @shrey_g]

Ok almost Shrey Gupta

[The Associated Press @AP]

BREAKING: Shrey Parikh wins the Scripps National Spelling Bee, beating Ishaan Gupta in a lightning-round tiebreaker. https://t.co/Gmw7USaE0Q

[Blake Burge @blakeaburge]

Underrated life advice: Make yourself easy to root for. Be kind. Be reliable. Celebrate other people’s wins. Work hard without complaining. Carry good energy into rooms. You'll be shocked by how many doors open for you by making life better for others.

[Similarweb @Similarweb]

Gen AI traffic share update

Main takeaways:
→ Claude and Gemini continue to grow.
→ ChatGPT moves closer to the 50% mark.

🗓️ 12 months ago:
ChatGPT: 77.6%
Gemini: 7.27%
DeepSeek: 6.01%
Grok: 3.17%
Perplexity: 1.75%
Copilot: 1.56%
Claude: 1.37%

🗓️ 6 months ago:
ChatGPT: 69.5%
Gemini: 15.9%
DeepSeek: 4.06%
Grok: 3.31%
Perplexity: 2.22%
Claude: 2.12%
Copilot: 1.97%

🗓️ 3 months ago:
ChatGPT: 61.2%
Gemini: 23.9%
Grok: 3.94%
DeepSeek: 3.09%
Claude: 3.29%
Copilot: 1.87%
Perplexity: 1.74%

🗓️ 1 month ago:
ChatGPT: 53.7%
Gemini: 26.7%
Claude: 7.95%
DeepSeek: 3.97%
Grok: 3.20%
Copilot: 1.98%
Perplexity: 1.50%

[kache @yacineMTB]

you can outsource your thinking
but you cannot outsource your understanding

[Shrey @shrey_g]

Earthquake night

[Shrey @shrey_g]

@KendraEBarnett @dseetharaman @Adweek Thanks. Pretty odd ad choice then.

[Shrey @shrey_g]

@antoniogm @signulll The experiment end was for integrated checkout not for Ads I believe

[Shrey @shrey_g]

I scored 187.5 on Pointed. Play at https://t.co/RtVvXsFIW7 so we can compare!

[Axios @axios]

The 2025 news cycle, in 1 chart https://t.co/PNSk42HvU3

[Shrey @shrey_g]

https://t.co/8OIAJEfzJg

[Andrej Karpathy @karpathy]

Happy weekend to those who celebrate

[Nano Banana 2 Lite @NanoBanana]

Prompt: Present a clear, 45° top-down isometric miniature 3D cartoon scene of [CITY], featuring its most iconic landmarks and architectural elements. Use soft, refined textures with realistic PBR materials and gentle, lifelike lighting and shadows. Integrate the current weather conditions directly into the city environment to create an immersive atmospheric mood.
Use a clean, minimalistic composition with a soft, solid-colored background.

At the top-center, place the title “[CITY]” in large bold text, a prominent weather icon beneath it, then the date (small text) and temperature (medium text). All text must be centered with consistent spacing, and may subtly overlap the tops of the buildings.

[Shrey @shrey_g]

@umachingunde @pitdesi I have some of those, one for Peloton. I think it is an invite for someone to connect their Peloton/Zillow account and then maybe even go there. So increasing commercial action acceptance before ads.

[Shrey @shrey_g]

@umachingunde @pitdesi Thanks for sharing. Yes, organic for now but there is a straight line here on how to insert ads to this layout.

[Shrey @shrey_g]

@umachingunde @pitdesi Interesting. Is this in chatGPT? Do you have a screenshot?

[Shrey @shrey_g]

Rome wasn't built in a day, same for @GoogleAI :)
Great talk by @JeffDean . It isn't easily visible outside but many breakthroughs (language models like BERT or scaling for DL) that get attention via LLMs have been used for Search, Ads, YT for a while now

https://t.co/jxvLgidHrR

[Shrey @shrey_g]

@benhylak 3) Basically in pure ML model world, we had offline metrics (PR-AUC, bias etc) and you could actually find sometimes A/B metrics (revenue because you show more ads) improving on a worse model. So you need some offline check in this case too.

[Shrey @shrey_g]

@benhylak 2) Causality: Say you A/B test says you new candidate is poor? What do you, you are stuck? What is a quality problem or some UX change? Evals check can help you measure quality before going to prod and limit that factor.