Parth Mishra's Avatar

Parth Mishra

@parth.dev

A cog in the Machine. ML Engineer @ Accenture. πŸ“Brooklyn, NY

211
Followers
319
Following
252
Posts
26.04.2023
Joined
Posts Following

Latest posts by Parth Mishra @parth.dev

Post image

THE FIRST AFRICAN AMERICAN *AND* FIRST WOMAN TO BE ELECTED NEW YORK ATTORNEY GENERAL PULLED UP TO TALK BROOKLYN, SUING TRUMP, AND SAVING PRESTON HIGH SCHOOL.

14.07.2025 21:48 πŸ‘ 10 πŸ” 2 πŸ’¬ 1 πŸ“Œ 0

I can't read manga because no matter how hard I try, I can only hear the characters voices as the worst 4Kids dub

09.07.2025 17:56 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Time to globalize the tostada

25.06.2025 14:16 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I would enjoy summer way more if it was socially acceptable to be wetter than a sloppy steak

24.06.2025 21:47 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Video thumbnail

On Friday night, we walked the length of Manhattan, from Inwood Hill to Battery Park.

New Yorkers deserve a Mayor they can see, hear, even yell at. The city is in the streets.

23.06.2025 15:06 πŸ‘ 14682 πŸ” 2553 πŸ’¬ 272 πŸ“Œ 744

I've literally never seen nor talked to a Cuomo supporter in real life while living in Brooklyn, so can someone tell me what is the consensus response for why they support a sex pest? Do they acknowledge this at all or...?

20.06.2025 16:05 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
A graphic of Bernie and Zohran smiling. Text reads β€œBERNIE ENDORSES ZOHRAN.”

A graphic of Bernie and Zohran smiling. Text reads β€œBERNIE ENDORSES ZOHRAN.”

As for so many across this country, Bernie Sanders has been the single most influential political figure in my life.

As Mayor, I will strive to live up to his example by fighting for the working class every day and hopefully make Brooklyn’s own proud.

17.06.2025 11:48 πŸ‘ 4198 πŸ” 541 πŸ’¬ 40 πŸ“Œ 77
Preview
The lethal trifecta for AI agents Plus reviews of two new papers about prompt injection, and Anthropic's tips on building multi-agent LLM systems

Sent out a new edition of my email newsletter (which is very genuinely just recent from my blog reformatted into newsletter form) and it turns out it's basically ALL prompt injection this week - two detailed paper reviews and my piece on the lethal trifecta simonw.substack.com/p/the-lethal...

17.06.2025 15:30 πŸ‘ 58 πŸ” 7 πŸ’¬ 3 πŸ“Œ 2

Seething at the audacity of β€œ#!/bin/bash”

How dare u assume I have bash and that it’s there

03.06.2025 23:25 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

anything by Terry Pratchet

28.05.2025 08:49 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Is this even a hot take? Regardless, I completely agree.

Just talking to a candidate for 30-60 min should really give you enough signal. Anything more is a waste of time for both parties involved.

13.02.2025 22:13 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

I'm gonna die mad at the fact that we call non-prescription medicine, the kind you can get anywhere in the store, "over the counter." But the medicine that the pharmacist literally has to hand me over a counter? That stuff is we classify as "not available over the counter." Fuck you!!

03.02.2025 17:04 πŸ‘ 1431 πŸ” 85 πŸ’¬ 32 πŸ“Œ 2
Preview
Revolutionary Suicide : Free Download, Borrow, and Streaming : Internet Archive BPP Huey Newton Revolutionary Suicide

Reading on its own is not inherently an act of resistance, but it becomes one if you allow it to inspire you to action.
archive.org/details/rshn...

02.02.2025 18:41 πŸ‘ 125 πŸ” 14 πŸ’¬ 2 πŸ“Œ 1

The problem with praise of DeepSeek is that the efficiency of LLMs is not the sole issue.

Outside of some narrow contexts, they lack any meaningful uses that aren't creatively and/or morally bankrupt. Their continued proliferation due to efficiency gains is ultimately a win for this grift

27.01.2025 17:33 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

the worst is AI slopologists who try to sell you on the fact that LLMs are just "helping" people write

27.01.2025 17:09 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

ive 10x'd my chip consumption after finding out you can use chopsticks to eat them

27.01.2025 17:07 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

in retrospect the e/acc movement is kind of hilarious. what do you mean we should keep accelerating technology? it’s already accelerating faster than it ever has in human history. e/acc was just a proponent of the status quo

26.01.2025 12:08 πŸ‘ 16 πŸ” 2 πŸ’¬ 1 πŸ“Œ 0

Why create a new term? It’s still just enshitification. The term is inclusive of any all forms of product degradation as a result of chasing profit and growth.

17.01.2025 23:14 πŸ‘ 3 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

Oh no! the poor peons who need to be employed in order to have health insurance aren’t taking the time to craft a bullshit story about how much they want to generate shareholder value with AI.

What are you going to do?!

17.01.2025 23:02 πŸ‘ 6 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Preview
Want to Defend Immigrant Workers in Your Contract? Here Are Some Suggestions. The following language was compiled from a series of unions and labor activists. It is intended as a resource for workers looking to include pro-immigrant provisions in their collective bargaining agr...

@nataschaelena.bsky.social and I compiled a repository of collective bargaining agreement language that defends immigrant workers, as well as resources on sanctuary resolutions and workers’ rights. labornotes.org/blogs/2025/0...

16.01.2025 17:02 πŸ‘ 34 πŸ” 17 πŸ’¬ 1 πŸ“Œ 2
Honeycomb - Staff Machine Learning Engineer What We’re Building Honeycomb is the observability platform for teams who manage software that matters. Send any data to our one-of-a-kind data store, solve problems with all the relevant context, and...

hey ML gang, got a cool job for ya. remote US.

jobs.lever.co/honeycomb/12...

15.01.2025 00:32 πŸ‘ 20 πŸ” 7 πŸ’¬ 0 πŸ“Œ 0

Why do you need to remove them?

Even if there were a way to remove them, I wouldn’t bother since just ignoring them is functionally the same provided they don’t spam you.

14.01.2025 20:07 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

There’s plenty of techno optimists e/acc blue check chuds left on Twitter.

They can stay there for all I care.

28.12.2024 20:25 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

This a weird straw man. The work life balance discourse is in opposition the to *expectation* that you need to be coding on nights and weekends.

If you want to code your own nights and weekends away, literally no one is stopping you.

22.12.2024 22:10 πŸ‘ 4 πŸ” 1 πŸ’¬ 1 πŸ“Œ 0

Ooph bud that’s a yikes from me pal

19.12.2024 22:46 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I don't get technical podcasts, how you retain that information audibly is beyond me, so I don't bother.

18.12.2024 19:41 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

To be clear, the above is for getting started with AI engineering, but staying up to do date as a current one is like you mentioned, blogposts, communities on here, and looking at what open source projects on Github are doing.

18.12.2024 19:40 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

AI engineers need some more cross disciplinary skills as well so an O'Reilly subscription has been invaluable for books on SWE, DevOps, MLOps, and ML in general. Although YouTube has been becoming a good source lately (e.g. Karpathy vids).

18.12.2024 19:37 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

yeah chess will fit right in as an esport if the chat is any indication

18.12.2024 19:01 πŸ‘ 7 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0

Tbh the fact that the lore of Arcane is even somewhat coherent after having only the equivalent of fantasy super smash bros as the starting point is a testament to their creative writing skills

18.12.2024 00:35 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0