Pranav Reddy
@pranavreddyg
Followers
333
Following
1K
Statuses
50
Co-founder of @PipeshiftAI (YC S24)
San Francisco, CA
Joined January 2019
RT @karpathy: New 3h31m video on YouTube: "Deep Dive into LLMs like ChatGPT" This is a general audience deep dive into the Large Language…
0
3K
0
RT @ycombinator: Congrats to the team @PipeshiftAI (YC S24) on the $2.5M seed! Pipeshift gives teams control and flexibility over their A…
0
4
0
RT @ClementDelangue: Our science team has started working on fully reproducing and open-sourcing R1 including training data, training scrip…
0
549
0
bro is autistic
DeepSeek R1 when it is prompted to select a random number: Okay, the user wants me to think of a number between 1 and 10,000 but not tell them. Hmm, they mentioned not to overthink it, so maybe just pick a random number quickly. Let me see. Well, people often have favorite numbers or patterns. Maybe go with something like 7777? But that's four 7s, which might be too obvious. Alternatively, 1234? Too sequential. Maybe a number that's a common choice but not too predictable. Wait, I should just pick the first number that comes to mind. Let me try that. Closing my eyes and... 42? Wait, that's from Hitchhiker's Guide, but maybe they don't want a reference. Alternatively, 1004? That's within the range. Or 8675? No, that's from a song. Hmm, maybe 3141? Like the first four digits of pi. But 3141 is 3.141... multiplied by 1000. That's clever, but maybe too nerdy. Alternatively, 1984? But that's a book title. Maybe 2023, the current year? But they might not want a recent year. Or 1337, which is leet speak. But that's only 4 digits. Wait, 1337 is 1-3-3-7, which is 1337. Maybe that's too slangy. Alternatively, 5000, the midpoint. But that's too obvious. How about 777? No, that's three digits. Wait, the number needs to be between 1 and 10,000, so up to four digits. Let me just pick 42 again. Wait, that's too low. Maybe 420? But that's associated with cannabis culture. Not sure. Maybe 6969? That's a playful number but might be seen as immature. Alternatively, 1111? A repeating digit. Or 2222? Hmm. Maybe 1234 is too simple. Wait, the user said not to overthink, so just pick a number without second-guessing. Let me go with 7321. No reason, just random digits. But maybe that's too arbitrary. Alternatively, 500, but 0500 is 500. Wait, leading zeros don't count. So 500 is 500. Maybe 8888? That's considered lucky in some cultures. But again, maybe too obvious. Alright, I'm overcomplicating it. The user said not to overthink. Let me just pick the first number that comes to mind. Okay, 7777. Done.
0
0
2
🗿
@abacaj deepseek’s holding 幻方量化 is a quant company, many years already,super smart guys with top math background; happened to own a lot GPU for trading/mining purpose, and deepseek is their side project for squeezing those gpus
0
0
0
RT @ClementDelangue: Once again, an AI system is not "thinking", it's "processing", "running predictions",... just like Google or computers…
0
1K
0
RT @LucasAtkins7: Scarlett Johansson’s work on seq2seq was instrumental to getting ML where it is today.
0
61
0
RT @ycombinator: YC S24's @PipeshiftAI is the cloud platform for fine-tuning and inferencing open-source LLMs, helping teams get to product…
0
14
0
RT @ChrisJBakke: Our family dog just left the house and announced that he’s joining Anthropic for TC of $2.7M.
0
35
0
Time for some new rubber 🏎️ Xylem AI is now @PipeshiftAI! We're building the infrastructure for fine-tuning and inferencing open-source LLMs at high speeds and low latencies. Quick and easy deployments, so that you can focus on what matters most: building the best AI apps and agents. If you are considering switching from GPT/Claude/Gemini to open-source models, we'd love to help. Our beta access waitlist is now open. Join the waitlist:
0
1
16
RT @AIatMeta: Starting today, open source is leading the way. Introducing Llama 3.1: Our most capable models yet. Today we’re releasing a…
0
1K
0