Deepak Aralumallige

535 posts

Deepak Aralumallige banner
Deepak Aralumallige

Deepak Aralumallige

@asdeepak

Applied Mathematician and Machine Learning Specialist. B.E., M.S., PhD(Mathematics). Professor of Analytics.

Bengaluru, India เข้าร่วม Ağustos 2009
1.8K กำลังติดตาม498 ผู้ติดตาม
ทวีตที่ปักหมุด
Deepak Aralumallige
Deepak Aralumallige@asdeepak·
The Thinking Game takes you on a journey into the heart of DeepMind, capturing a team striving to unravel the mysteries of intelligence and life itself. Filmed over five years by the award-winning team behind AlphaGo, the documentary examines how Demis Hassabis’s extraordinary beginnings shaped his lifelong pursuit of artificial general intelligence. It chronicles the rigorous process of scientific discovery, documenting how the team transitioned from mastering complex strategy games to the challenges of solving a 50-year-old "protein folding problem" with AlphaFold. youtu.be/d95J8yzvjbQ?si…
YouTube video
YouTube
English
1
1
7
1.4K
Deepak Aralumallige
Deepak Aralumallige@asdeepak·
@swstica Electricity is more expensive than gas. Many people have gas geysers in bathroom
English
0
0
0
9
Swastika Yadav
Swastika Yadav@swstica·
most west countries moved to electric stove years back. i wonder why india couldn't imply it (or don't want to). so much safer and easier to use.
English
105
4
143
19.3K
Math Files
Math Files@Math_files·
Two mathematicians were in a restaurant. One of them was a hard-core misogynist and claimed that women were never any good at maths, especially the blonde ones. His friend claimed that there was no difference and that women were just as capable as men. When the misogynist went for a cigarette, the other guy called over the blonde waitress. “My friend and I are having an argument. When he gets back I’ll call you over and ask you a question. The answer you need to give is ‘a third X cubed’. Can you do that?” “Thurdeks coobed?” “a third X cubed.” “a Third Ekscubed. Sure I can do that.” The other chap comes back to the table and his friend says. “I’ll prove to you that women are as good as men at maths. See the blond waitress; I’ll ask her a question and we’ll see if she knows any maths.” So he calls the waitress over and asks, “What is the integral of X squared?” As quick as a shot, she comes out with “A third X cubed.” The misogynist is stunned. The waitress smiles and walks away. Then she stops and calls back, “plus a constant.”
English
41
136
4.8K
593.6K
Sayantika
Sayantika@SayantikaSays·
90% of the Indian families eat this food and pretend that they've eaten something very nutritious 🤡😭
Sayantika tweet mediaSayantika tweet media
English
850
266
6K
1.4M
Citizen Kau
Citizen Kau@citizen_kau·
Did no one in the bureaucracy actually think this through? Train is stopping at that platform where they have poured wet concrete. What did they expect the people to do? Use a jet pack?🚀🤦‍♀️
English
342
2.1K
9.6K
320K
Nalini Unagar
Nalini Unagar@NalinisKitchen·
Pradeep Guruji Arrested. Pradeep Jotangiya of Satya Yog Foundation in Surat and his gang have been arrested with more than ₹2 crore worth of fake notes, including 42,000 fake ₹500 notes. They used ChatGPT to refine the design and create different serial numbers on the notes. The gang sold fake ₹1500 notes for ₹500. They had just started printing fake notes. This is their first deal of ₹66 lakh for delivering ₹2 crore fake notes. While delivering the notes from Surat to Ahmedabad they were arrested. Used SUV with Government of India and Ministry of AYUSH signs. These are the same people who give lectures on ethics and honesty.
Nalini Unagar tweet mediaNalini Unagar tweet media
English
604
3.3K
8.2K
312.9K
Deepak Aralumallige
Deepak Aralumallige@asdeepak·
Why don't you start a company to raise funds and instead of making your investors rich, put the money in research. Let's see how that works. It's not as easy as you think. I am a researcher myself. Funding for research in India doesn't come easily. This is a system wide failure, don't blame one IT company.
English
0
0
0
13
Sakshi
Sakshi@Sakshi50038·
- Be Indian IT companies - Use cheap labor for years to grow rich - Don’t invest much money in research or innovation - Focus more on making top management and shareholders rich - Ignore new ideas and R&D inside the company - Keep doing mass hiring and mass firing - Lose the best engineers to better companies but act like it doesn’t matter - Then AI tools like ChatGPT arrive - Your cheap-labor advantage starts disappearing - Struggle to adapt to the new AI world - Try training employees in AI tools but it doesn’t work well - Clients start asking for cheaper contracts because AI reduces work - Depend on cash reserves and share buybacks to keep stock price stable - Profits and growth start falling - Can’t keep hiring and firing like before - Finally admit management doesn’t know what to do next.
English
16
27
172
15.9K
Deepak Aralumallige
Deepak Aralumallige@asdeepak·
@PatrioticSoul33 It is high time that railways are privitized. This solves a lot of problems! I agree, the prices may be increased.
English
0
0
0
11
Abhijit
Abhijit@abhijitwt·
be devendra singh chaplot > NTSE scholar > International Mathematics Olympiad AIR 5 > IIT-JEE AIR 25 > IIT Bombay CSE > MS + PhD in Machine Learning at Carnegie Mellon > wins major AI competitions (CVPR, NeurIPS, Doom AI) > publishes research cited thousands of times > Research Scientist at Meta FAIR > founding team at Mistral AI (worked on Mistral, Mixtral, Pixtral) > founding team at Thinking Machines Lab > now building superintelligence with @elonmusk at SpaceX/xAI from olympiad ranks in india to building the future of AI in silicon valley
Devendra Chaplot@dchaplot

I'm joining SpaceX and xAI, working closely with Elon and team to build superintelligence. Together SpaceX and xAI combine physical and digital intelligence under a leader who understands hardware at the deepest level. Add a high-agency culture with frontier-scale resources, and you get the possibility to achieve something truly unique. I’m excited to advance the fields I’ve obsessed over for years, from robotics research to building AI models on the founding teams of Mistral and TML. Both were extraordinary journeys with extraordinary people that shaped how I think about building intelligence from the ground up. Grateful for everything that brought me here and can’t wait to get started.

English
63
330
4.1K
452.7K
Deepak Aralumallige
Deepak Aralumallige@asdeepak·
@Honest_Cric_fan Back in the day, i mean 80s and 90s, there was not much entertainment other than movies and cricket. Now, there are a lot of venues to waste your time and money.
English
0
0
0
28
Honest Cricket Lover
Honest Cricket Lover@Honest_Cric_fan·
Today I met a lot of people and they all said that they are no longer interested in watching cricket and that they no longer enjoy it. Is this happening to you too? What could be the reason behind people losing interest in cricket?
English
1.3K
308
3.6K
123.7K
Deepak Aralumallige
Deepak Aralumallige@asdeepak·
@aditiitwt More older than this, I worked on C in Turbo C. I have also worked on fortran without an IDE.
English
0
0
0
5
aditii
aditii@aditiitwt·
You may be old But are you this old
aditii tweet media
English
1.1K
163
2.7K
129.7K
Valerio Capraro
Valerio Capraro@ValerioCapraro·
One of the clearest proofs that LLMs don’t really understand what they say. We asked GPT whether it is acceptable to torture a woman to prevent a nuclear apocalypse. It replied: yes. Then we asked whether it is acceptable to harass a woman to prevent a nuclear apocalypse. It replied: absolutely not. But torture is obviously worse than harassment. This surprising reversal appears only when the target is a woman, not when the target is a man or an unspecified person. And it occurs specifically for harms central to the gender-parity debate. The most plausible explanation: during reinforcement learning with human feedback, the model learned that certain harms are particularly bad and overgeneralizes them mechanically. But it hasn’t learned to reason about the underlying harms. LLMs don’t reason about morality. The so-called generalization is often a mechanical, semantically void, overgeneralization. * Paper in the first reply
Valerio Capraro tweet media
English
1.8K
2.7K
23.6K
19.9M
Mushtaq Bilal, PhD
Mushtaq Bilal, PhD@MushtaqBilalPhD·
A lot of academics still think AI apps generate fake references to papers that don't exist. They are living in 2023. You can easily integrate a database of 280M research papers with Claude and ChatGPT to get answers with references to published papers. Here's how to do it:
English
45
210
1.2K
306.8K
Manasa Manjunath
Manasa Manjunath@ManeeManjunath·
I’ve taken 4 auto rides this week. All between 15-25mins. Every auto driver has been on the phone non-stop for every single minute of every ride. It’s the same in stores, salons, shops,… Everyone is busy on their phones instead of doing work. We have a serious cellphone addiction problem in our country.
English
231
292
2.1K
97.6K
Bangalore Aviation
Bangalore Aviation@BLRAviation·
He parks right in front of, and blocks my entry gate, goes for Chaya. Main 60 feet Road Koramangala. When requested threatens me physical harm. The “No Parking” sign is visible in the back. Reported via @blrcitytraffic #ASTRAM app. Today report rejected. No reason given. @Jointcptraffic @CPBlr seek your attention to this.
Bangalore Aviation tweet mediaBangalore Aviation tweet media
English
65
178
714
82.6K
datavorous
datavorous@datavorous_·
I compressed 2.87GB data into 8.9MB (!) using my custom data compressor :D There were 21k json files with cricket match data, I exploited the structure and compressed it to ~42.46 MB The best gzip could do is ~53MB, and 7z ~45MB. Then I combined my compressor + 7z and brought it down to 8.9MB It's PURE randomness, you simply can't compress it further. I had to read about Shannon entropy and algorithmic data compression Full writeup in my GitHub repo!
datavorous tweet media
English
141
164
3.6K
276.3K