sidd @siddgood

ml infra swe @ baton 🚚 • currently fine-tuning my brain 🧠 siddgood.bearblog.dev San Francisco, CA Joined January 2026

Tweets

17
Followers

18
Following

293
Likes

2K

sidd @siddgood

3 days ago

@DylanCalluy Omggg where’s this?

0 0 0 45 0

View Details

@simonw Yeah…I’ve been noticing the quality of response is not as a great. It’s a bummer because I often try to use it to entertain ideas, but then quickly find it degrading. I believe it has to do with the fact that the voice model doesn’t have the advanced toolset use capabilities.

0 0 0 795 0

View Details

sidd @siddgood

5 months ago

Project Hail Mary 👎

0 0 0 87 0

View Details

sidd @siddgood

5 months ago

x.com/i/article/2032…

0 1 1 121 1

View Details

sidd @siddgood

6 months ago

@SarahChieng @allenpark @zhennydez I’d love to join!

0 0 0 838 0

View Details

sidd @siddgood

6 months ago

Why did I just end up watching this in full?

Coinbase 🛡️ @coinbase

6 months ago

An overview of our Q4 and full year 2025 financial results. With something extra to keep you focused.

827 429 6K 4.1M 1K

0 0 0 134 0

View Details

sidd @siddgood

7 months ago

Morning run!

0 0 1 61 0

View Details

sidd @siddgood

7 months ago

“Token usage” is the new line item, similar to cloud costs and ad spend. Current companies and startups might have large token usage, indicating a proxy for scale. Eventually, they will be judged by the value created relative to token spend. I see a future where entire teams, PMs, and maybe new companies focused entirely on measuring and optimizing the value-to-token-usage ratio.

0 0 1 50 0

View Details

aditya addepalli @adiaddxyz

7 months ago

read this and immediately had to build it. 5-layer memory architecture: → sensory buffer (attention scoring, noise filtering) → short-term memory (entity extraction, clustering) → long-term memory (consolidation + decay curves) → core memory (identity via reflection) → background "sleep" processes the future of memory is cognition, not glorified cache storages. open sourced github link below ⬇︎ stack: - @convex (db, agent, action cache, rate limiter, workflow) - @AnthropicAI haiku & @OpenAI embeddings - @aisdk, @base_ui, @tailwindcss, @shadcn, @tan_stack start, @colinhacks's zod, and @haydenbleasel streamdown

Manthan Gupta @manthanguptaa

7 months ago

x.com/i/article/2006…

43 47 512 85K 763

1 1 5 314 0

View Details

sidd @siddgood

7 months ago

a stroll on a cloudy sf morning gets you feeing some sort of way, especially after your hooping plans get cancelled due to rain…

0 0 3 39 0

View Details

sidd @siddgood

7 months ago

recently have been watching @karpathy’s “Zero to Hero” series to better understand the architecture of LLMs from the ground up! even though i have an ml background, it’s so refreshing to learn and re-learn certain basics that i didn’t quite grasp back in college!

0 0 0 31 0

View Details

sidd @siddgood

7 months ago

i’m going to start a new series here called “things i vibe with.” it’s a culmination of quotes and phrases from people, books, internet, etc. that i found valuable and “vibed” with. i’ve been compiling this list for some time now, but thought it might be fun to share! 🤙

0 0 1 36 0

View Details

sidd @siddgood

7 months ago

@bryan_johnson Getting 8 hours of sleep per night with a consistent bedtime (and eating my last meal at least 3 hours before).

0 0 0 3 0

View Details

sidd @siddgood

7 months ago

happy new year 🥳! starting a fresh account this year to focus on my goal of building a better personal brand and engaging more with vibrant tech community here in sf (and other cities too)! excited for a great year of learning and connecting with others!