agnel

technologist | cloud ☁️ and security | interested in indic languages, history, epigraphy, and tech | views are personal

[agnel @_agnel_]

Very informative interview with VP of AI Software at Tesla, Ashok Elluswamy @aelluswamy by Gopinath in Tanglish.

https://t.co/vRyew8Bg2M

[agnel @_agnel_]

@RamprashanthVe1 😅

[agnel @_agnel_]

@RamprashanthVe1 What’s the etymology for chal? cel- seems like in most Dr lgs.

[agnel @_agnel_]

@cobbaltt @David_Erschler Aren't you supposed to use your father’s name as your surname in your passport?

[agnel @_agnel_]

@cobbaltt @a_a_viswanath Which dialect loosing the diphthong?

[agnel @_agnel_]

@cobbaltt Are you pursuing your PhD now?

[agnel @_agnel_]

On a recent visit to Thiruvarur Thiyagaraja Swamy temple, I found this mural on the ceiling of the corridor. It probably depicts the goddess with two angels holding a child.

Interesting, wonder any background story to these. Could be a Nayaka period mural?

[agnel @_agnel_]

@cobbaltt Welcome back!

[Aravind Srinivas @AravSrinivas]

And beating DeepSeek R1 on all benchmarks with rigor will mean I will invest $10mm more.

[agnel @_agnel_]

“The essence of human intelligence might be language, and human thought could essentially be a linguistic process. What you think of as ‘thinking’ might actually be your brain weaving language” - Liang Wenfeng, in an interview from 2023.

[Andrej Karpathy @karpathy]

DeepSeek (Chinese AI co) making it look easy today with an open weights release of a frontier-grade LLM trained on a joke of a budget (2048 GPUs for 2 months, $6M).

For reference, this level of capability is supposed to require clusters of closer to 16K GPUs, the ones being brought up today are more around 100K GPUs. E.g. Llama 3 405B used 30.8M GPU-hours, while DeepSeek-V3 looks to be a stronger model at only 2.8M GPU-hours (~11X less compute). If the model also passes vibe checks (e.g. LLM arena rankings are ongoing, my few quick tests went well so far) it will be a highly impressive display of research and engineering under resource constraints.

Does this mean you don't need large GPU clusters for frontier LLMs? No but you have to ensure that you're not wasteful with what you have, and this looks like a nice demonstration that there's still a lot to get through with both data and algorithms.

Very nice & detailed tech report too, reading through.

[agnel @_agnel_]

@yajnadevam Congrats! Good luck 👍🏽

[Sam Altman @sama]

🎄🎅starting tomorrow at 10 am pacific, we are doing 12 days of openai.

each weekday, we will have a livestream with a launch or demo, some big ones and some stocking stuffers.

we’ve got some great stuff to share, hope you enjoy! merry christmas.

[Arpit Bhayani @arpit_bhayani]

Why are managers even needed if they are not coding? are they supposed to code? why are they paid so much? And more importantly, should you become a manager or not? This is pretty common to wonder especially as early engineers. Here are a few pointers from my experience

The bulk of the work a manager puts in is abstract and super difficult to quantify. This gives an illusion that the manager is "not doing anything". To ensure a significant impact and the outcome of the team, a manager needs to do at least three things really well

- prioritize the right projects
- track the right metrics
- reward the right people

Given managers hold the authority on the execution and are accountable to the team, they need to be right a lot. This sounds easy, but it is not.

Managers need to have the right technical acumen, some technical know-how, and the right gut feeling to make the right decisions across design, engineering, product, and business.

When I was a manager I was under constant pressure to deliver things, to make sure the right projects were picked up, and to ensure all engineers (and managers) had enough impactful work on their plate and were getting recognized and rewarded.

I could not cut some slack because the career growth of my team was my responsibility and making sure it happened was all on me. So, if you are in a dilemma, you should become a manager if

- you prefer mentoring people
- you want to make sure people grow
- you are patient, observant, and a good listener
- you can get people excited to chase a goal
- you communicate well
- you are organized and have an urge to keep things in check

Let me get this straight, whether managers should code or not is either a personal or organizational preference. In the end, what matters is the impact and outcome the person is making.

But one thing I can tell, if you choose to become a manager, be ready to spend a ton of time with others :)

ps: enroll in my sys design dec cohort and be part of amazing brainstorming sessions that will help you build the right intuition to design and understand any and every system. only 13 seats left.

https://t.co/nnhcj5gQzH

#CareerGrowth #AsliEngineering

[Suresh Kolichala @SureshKolichala]

Did Brahui retain the archaic form with frontvowels for the word for stomach in PDr? This was possibly borrowed by early Prakrits and hence pēṭ- related words in several IA-langs. In PPDr langs, frontvowels between labials and retroflex tend to acquire backness hence poṭṭa.1/n

[agnel @_agnel_]

இனிய பொங்கல் நல்வாழ்த்துகள்!

Cover of Vellimani, Pongal special weekly from 1948.

[agnel @_agnel_]

@holatombola Generic name he says

[agnel @_agnel_]

Wiktionary says my name comes from Friulian agnel 'lamb,' derived from the Latin 'agnellus' -> agnus + -ellus (diminutive suffix).

The Latin 'agnus' eventually comes from Proto-Italic *agʷnos < PIE *h₂egʷnós 'lamb.' Interesting, it doesn't have any IIr reflexes.

[The Layman's Linguist @LaymansLinguist]

What does your given/chosen name mean etymologically?

Along with Caitlin/Katherine/etc, <Kathleen> originally comes from the ancient Greek name <Ἑκατερινη> "Hekaterine" as in Hecate, the three-faced goddess of sorcery, crossroads, the moon, poisonous plants, and necromancy.

[agnel @_agnel_]

@avzaagzonunaada ☹️