Uddy
@itsuddy.bsky.social
From being bullied as a kid to leading 8 digits agencies and enterprises and now helping others scale with my experience.
Find more about me: uddeshyaagrawal.com
Find more about me: uddeshyaagrawal.com
Holy shit DeepSeek did it again!! 🔥
Deepseek V3.2-Speciale just dropped.
- Rivals Gemini 3.0 Pro
- Gold medal IMO, ICPC, IOI 2025
- Reasoning-first. Built for agents.
- Fraction of Google's $100B budget
At what point does Silicon Valley admit they're getting outbuilt? 🤔
Deepseek V3.2-Speciale just dropped.
- Rivals Gemini 3.0 Pro
- Gold medal IMO, ICPC, IOI 2025
- Reasoning-first. Built for agents.
- Fraction of Google's $100B budget
At what point does Silicon Valley admit they're getting outbuilt? 🤔
December 1, 2025 at 12:05 PM
Holy shit DeepSeek did it again!! 🔥
Deepseek V3.2-Speciale just dropped.
- Rivals Gemini 3.0 Pro
- Gold medal IMO, ICPC, IOI 2025
- Reasoning-first. Built for agents.
- Fraction of Google's $100B budget
At what point does Silicon Valley admit they're getting outbuilt? 🤔
Deepseek V3.2-Speciale just dropped.
- Rivals Gemini 3.0 Pro
- Gold medal IMO, ICPC, IOI 2025
- Reasoning-first. Built for agents.
- Fraction of Google's $100B budget
At what point does Silicon Valley admit they're getting outbuilt? 🤔
GPT-5.1 dropped 5 minutes ago.
Ran it through my benchmark dataset immediately.
It's not just better. It's significantly better.
Beats Grok. Beats Claude.
Instruction following is near-perfect. Adaptive thinking actually works.
This isn't incremental improvement.
This is a leap toward AGI.
Ran it through my benchmark dataset immediately.
It's not just better. It's significantly better.
Beats Grok. Beats Claude.
Instruction following is near-perfect. Adaptive thinking actually works.
This isn't incremental improvement.
This is a leap toward AGI.
November 12, 2025 at 7:53 PM
GPT-5.1 dropped 5 minutes ago.
Ran it through my benchmark dataset immediately.
It's not just better. It's significantly better.
Beats Grok. Beats Claude.
Instruction following is near-perfect. Adaptive thinking actually works.
This isn't incremental improvement.
This is a leap toward AGI.
Ran it through my benchmark dataset immediately.
It's not just better. It's significantly better.
Beats Grok. Beats Claude.
Instruction following is near-perfect. Adaptive thinking actually works.
This isn't incremental improvement.
This is a leap toward AGI.