Hesam
@itsHesamSheikh
Followers
1,419
Following
397
Media
653
Statuses
7,421
ML research tweets; Writing AI blog posts @TDataScience to help you and me learn more...
Joined October 2023
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
Jack Smith
• 272933 Tweets
Arlington
• 247971 Tweets
A'ja
• 230597 Tweets
Juan Izquierdo
• 150837 Tweets
lana
• 47330 Tweets
Dana Bash
• 37742 Tweets
Descanse
• 31607 Tweets
Aces
• 31029 Tweets
ムアラニ
• 29936 Tweets
野人転生
• 26496 Tweets
#WWENXT
• 25857 Tweets
最高の日
• 23295 Tweets
Q31 U STEAL MY HEART
• 21387 Tweets
Marley
• 20911 Tweets
ズゴック
• 19516 Tweets
Positive Vibes
• 17934 Tweets
Unna
• 17066 Tweets
Saint Dr MSG Insan
• 13458 Tweets
#nasne12周年
• 12491 Tweets
Nacional-URU
• 12229 Tweets
Soulful Living
• 11314 Tweets
コンクリ事件
• 11094 Tweets
Dr. Phil
• 10220 Tweets
Pinned Tweet
🚀 An easy-to-understand introduction to AI Agents:
1
1
6
you're a Programmer,
you hit an issue,
you Google it,
every link is purple.
198
265
2K
i accidentally deleted a day's work of code :) . I hate git
198
7
376
@svpino
Im not sure Jupyter notebook is a good choice for software development? 🤔
It’s best for small prototyping
9
3
370
wasn't Wix supposed to put web developers out of work?
what happened there?
39
6
114
Do you know what the difference between Programmer and Developer is?
71
3
99
@svpino
An example of when you take political bias out of data and treat the data as an unbiased and neutral data analyst 👌
1
6
81
Finally got the blue tick!
All that remains is to find 17 more people to connect with.
#programmer
s let's connect!
32
3
66
@bindureddy
but you still have to remember, that after MONTHS others are catching up to OpenAI products. And the same will most likely happen for GPT-5.
OpenAI is surviving because of its innovation, not of offering cheaper options.
10
0
65
Tomorrow is my first work day at my new job ✨ . AI pays the bills
14
3
64
Hey X! I'm looking to
#Connect
with:
🌐 Full-Stack Developers
💻 Backend Developers
🚀 AI/ML Engineer
📊 Data Analysts/Scientists
🔨 Data Engineers
🧲 BI Developers
💼 Business Analysts
🛠️ DevOps Engineers
Drop a 👋
#letsconnect
#buildinpublic
94
2
64
@karpathy
If a LLM is able to convert Python to an adequate C file,
It’s safe to say we have AGI 💪🏻
10
1
51
@nixcraft
Consuming free content but refusing to thank the creator by watching a few seconds of ads <<<
70
0
46
The number of unfollows is surprising, all from accounts I haven't followed back.
I just follow whoever posts authentic and interesting tweets.
Just stop this childish game of trading follows.
16
0
45
Do i have to post that guys with the "developers let's connect" sign?
I just want to connect with great people. I'll be thankful if you help.
}:‑)
20
4
39
@karpathy
It’s so surprising you mentioned that in the beginning the C version of the code is slower than the Python implementation of it
This is a proof that when it comes to speed, language is not the only factor,
EVEN comparing C with Python
EVEN with training an LLM
And one might
11
1
39
@ylecun
interesting we cannot even come up with a standard description of what "human-level intelligence" even is
4
0
38
Why do startups fail?
if you don't know, then don't start your own company or project...
30
2
36
Which kind of software engineer will never succeed in their career?
31
0
34
if you could marry a language, what would it be?
i'd marry Python, it's easy-going
26
0
25
2023 is the year i met this guy
\ (•◡•) /
\ /
——
| |
|_ |_
And i muted him forever
9
0
32
You can be the best programmer, and the worst guy with social media.
You can have thousands of followers and still be the worst programmer.
13
3
30
✨ If you're a Programmer or Developer who wants to know more about Software Engineering and how to handle
#Data
in your application, bookmark this post.
I will share some of the interesting points I learn from this book as I read along and add new posts in this thread.
4
1
30
@engineers_feed
Maybe we are just too boring for species with the ability of extraterrestrial travel to bother
0
0
31
With a little push, I can have 200 connections!
It's going faster than expected...
12
0
29
@engineers_feed
And if you look at the buerocrocy and political system if any poor country, you see the very same patterns happening
2
4
31
how do you handle the FOMO of everything that is going on in AI?
26
0
29
techies are making a good buck,
but what is the next money maker, 20 years from now?
18
1
25
@burkov
I get it, but still, what's the reason for open-sourcing it? the precision is good enough to even be proprietary.
8
1
28
What should programmers learn for the next 10 years in order to stay in the market?
64
1
27
@svpino
+ Here are the next steps:
- learn numpy
- learn Tensorflow/Pytorch/Keras
- train a neural net to solve a vision classification task
0
5
26
a paper published in ELSEVIER starts with ChatGPTs words
It makes you think, does anyone even read these papers before publishing?
8
2
26
"CS grads have a higher chance of landing a job in tech than their peers", do you agree?
19
1
26
@svpino
This might be the most clever approach I have seen for any ML problems
All the websites and tutorials give you the same tips that take time and are not that effective, but this seems very promising
I’ll be sure to test it out 👌
Is this from your experience or from a book?
1
0
26
Google's demo of Gemini is FAKE!
If you have seen the video you noticed the fast response of Gemini to a video and accurate answers.
Well,
It's fake!
Google captured a video,
Then, used still images from it,
And then prompted Gemini using TEXT.
It's a shame.
#GeminiAI
8
1
25
I don't get Ghostwriting. Why would I PAY someone else to pretend to be ME?
14
0
23
Two of the best shows ever, are about tech.
Do you know any other great tech movies/shows?
20
1
24
@engineers_feed
This is because of how the chat is designed compared to a messenger such as Whatsapp.
You’re using a chat designed by the company itself and so this gives them access to your writings at any given time.
1
0
24
Twitter algorithm doesn't favor sharing information.
If favors memes and let's connect posts.
7
1
21
@bindureddy
I see a pattern that 99% of the people super excited about AI replacing programming are people who make money out of it.
2
0
24
@Prathkum
tell me you're a geek without telling me you're a geek LOL
but seriously, amazing 👏
2
0
23
I'm the only one in here who isn't trying to sell you anything :S
8
0
19
One month ago, I shared an article about Kolmogorov–Arnold Networks (KAN) on
@TDataScience
,
it's my most-read article to date and I'm so grateful for how well it was perceived. 🙏
KAN presents a fresh perspective on Neural Networks and suggests a possible alternative to
1
3
21
@svpino
the hype around Mojo started more than a year ago,
now with it being open-source we can actually see something happening ✨
0
0
23
The most underrated skill that nobody teaches, is HOW TO READ CODE. 📓
6
0
21
@engineers_feed
They have no idea what they’re doing, while they’re the only people who are supposed to
3
0
22
@stats_feed
It wouldn't make any difference
Anything the mankind created he would have created in many other points in history as well
Do you think if Atomic Bomb was not made, nobody else would have created it by now?
6
1
20
@KevinNaughtonJr
If C is so bad, why is the infrastructure of almost every server built on it?
2
1
21
🔥KANs: Kolmogorov–Arnold Networks are said to be a "new revolution" of machine learning or "game-changing".
But after a week, let's see if they deserve the hype:
3
4
22
Reasoning with LLM is Hard!
Large Language Models need help with generalized reasoning capabilities, and a key factor is how we prompt them.
📌 Traditional prompting methods such as Chain-of-Thought (CoT) or Tree-of-Thought (ToT) often require multiple assumptions or numerous
0
2
21
Python is SLOW,
But it's:
- the King of AI
- Much faster to develop
- Rich with Libraries you Need
The downside? It's SLOW.
Now imagine a programming language as sleek as Python but 68000x Faster.
This is what Mojo 🔥claims:
2
2
20
I really don't see how ChatGPT can replace an average programmer "at this point".
If you're worried about that, improve.
6
2
19
@akshay_pachaar
Clearly and well explained
It’s one of the tough things to get your head around but once you do then you see it’s not that hard
0
0
18
@engineers_feed
top 10 tech inventions of all time:
1.Electricity
2.Internet
3.Printing Press
4.Telephone
5.Transistors
7.Steam Engine
8.Antibiotics
9.Automobile
Phone
4
2
20
running ML models Locally is easy,
running in Production is HARD,
just like engineering where manufacturing is much harder than design, in AI it is also more challenging to run ML models at scale.
Here's Why:
- You need to serve 1000s of users
- You need to distribute
0
0
20