Akash Ashok

AI @Rippling | Machine Learning and Engineering

[Akash Ashok @thehellmaker]

Really impressive how close open weight models are (esp GLM 5.2) on our Rippling Bench (Evals for Rippling AI).

I am getting work on this because of all these cool folks @fullung @stanine @laks_srini

[Matt MacInnis @stanine]

New: we ran ~2100 scored runs on open-weight models and lab models and found some big surprises. Grok (@spacexai) was the value leader, and GLM (@Zai_org) is super close to the frontier. I guess this is why the 'labs' are nervous... https://t.co/BjpVf9QFk2

[Albert Strasheim @fullung]

Nice to see the @typesafeai sales team replying to emails at 6:33 PM on a Friday. Guess someone forgot to send them the memo about pacing the frontier... 😅

[san @saneord]

"we have sandboxed the agent"

the agent:

[Akash Ashok @thehellmaker]

@stanine Roko's Basilisk shall spare you 😊

[Akash Ashok @thehellmaker]

@elonmusk increase number of tools the model supports. It's cries saying doesn't support

Cached token pricins is too high ...

[Akash Ashok @thehellmaker]

@stanine Woohooo. Thanks @stanine @fullung.

[Parker Conrad @parkerconrad]

https://t.co/OPrGjFcocj

[Parker Conrad @parkerconrad]

We launched Rippling Data Cloud today - an all-in-one rebuild of the modern data stack, with AI deeply integrated throughout.

Why would you want an org-and-employee-centric data stack? Well, here’s how I used Rippling Data Cloud to help with token burn and cut AI slop. 1/

https://t.co/2rhp8x3KVn

[Akash Ashok @thehellmaker]

Been waiting for Wispr Flow on Android? Win stuff on this 🎰 waitlist: https://t.co/COmGySz5EN

[Akash Ashok @thehellmaker]

@Akshat_World @grok Is this correct?

[Akash Ashok @thehellmaker]

@lateinteraction This would be great

[Jonathan Gorard @getjonwithit]

Calling c the "speed of light" completely misses the point. Rather, c is the "spacetime exchange rate": how many units of space you can exchange for one unit of time.

In actuality, everything travels at the "speed of light", just not necessarily through space alone... (1/4)

[Hussein Nasser @hnasr]

Being humble is just another form of ego.

[Akash Ashok @thehellmaker]

Very good advice on reading research papers. Came across this other post by @abhi9u on the same topic which is written so we'll. It's lovely.

https://t.co/hkq0QWIsjU

[Arpit Bhayani @arpit_bhayani]

If you want to read papers, then read; simple. Don't let anyone dictate what you should or shouldn’t do.

Do not limit yourself to someone else's curation.

It might be overwhelming at first, but over time it will become much easier. Pro tip: try to prototype the core of the paper, and use LLMs if required.

Do not optimize for retention, focus on understanding it. You form the dots in your head and get your brain rewired with this new information is more important. Because when the time comes, you will recall the required things.

There is no roadmap. Start with the paper that interests you. This intrinsic motivation will help you overcome the difficult segments and convoluted concepts.

[Akash Ashok @thehellmaker]

@Jyotinder_Singh What's your view point on determinist simulation testing for DBs? I read about why it's important for databases here.
https://t.co/qNwgRvIYmc

[Vrushank Desai @vrushankdes]

i find it hilarious that Deepseek released pages of details about their parallelism/quantization/etc to stave off haters doubting their training efficiency and.....haters still can't believe how efficient their model is lmaoo

[Akash Ashok @thehellmaker]

@arpit_bhayani Fair

[big dele ✰ @deleszn_]

Too real😂

[avee @blueavee]

rust btw