AshutoshShrivastava
@ai_for_success
Followers
63K
Following
119K
Media
9K
Statuses
64K
Post about latest AI news, tools, tutorials and memes. Join my newsletter : https://t.co/jZDvYbKLa3
sponsorships š© or check my š
Joined January 2023
Analyzing vast amounts of market data for stock screening is incredibly time-consuming and overwhelming. FinBud eliminates this problem entirely. @finbudai is the most powerful AI agent built for stock screening that instantly processes millions of data points. - Screens
7
12
47
For every software engineer,Ā the job is going to shiftĀ from being an engineerĀ to being a CEO. You now have the toolsĀ to do so much more,Ā so I think that meansĀ you should aspire bigger. ~ OpenAI VP of Engineering Srinivas Narayanan
6
2
32
Large Language Models are improving at an exponential rate. If the pace continues until 2030, they will be able to complete, in hours, tasks that takes a human a month (167 working hours)
17
19
131
A Google researcher found that LLMs can embed hidden information and reasoning in their outputs raising concerns that we may soon lose the ability to fully monitor their internal behavior.
Can frontier models hide secret information and reasoning in their outputs?. We find early signs of steganographic capabilities in current frontier models, including Claude, GPT, and Gemini. š§µ
6
5
41
She couldāve just said no but she chose to humiliate him š„²
Thanks to Zuck, the new line is:. āsorry my salary is more than your entire Series - Bā
11
2
34
Remember the hype around PhD-level agents from OpenAI?. If that buzz was real, Iām 100% sure OpenAI wouldnāt mind losing a few more researchers to Zuck as they could just use those agents after all.
39
14
314
Cursor Pro unlimited plans details. "unlimited usage" was only for Auto and not all other models, which have at least $20 of included usage. Based on median token usage, the Pro plan currently covers about 225 Sonnet 4 requests, 550 Gemini requests, or 650 GPT 4.1 requests.
We recently updated our pricing, but missed the mark. We're refunding affected customers and clarifying how our pricing works.
7
3
24
On one side, thereās AI.On the other, thereās Soham Parekh. Even the co-founder of Huggingface said he applied there. Meanwhile you, who can't even get your resume selected.
Soham Parekh 50 Index š¤£. Maybe A16Z and other VCs should start tracking the companies Soham applied to, worked at, or did trial periods with. Some big names on that list.
4
3
40
If you donāt use AI, someone else will and theyāll outpace you.
18
3
102
MedSAM: Segment Anything in Medical Images.Open source has made a huge difference in the past and will always do.
š Huge milestone: MedSAM has officially surpassed 2,000 citations! š . MedSAM was the first foundation model for medical imaging! When we open-sourced MedSAM, our goal was to empower the entire medical imaging communityāand seeing it become a global standard is beyond
2
6
57
RT @bookwormengr: BRING-YOUR-OWN-BRAIN:.=========.Thanks you @kyutai_labs team, you are very cuteš I tried it, it is so refreshing and smarā¦.
0
3
0
The xAI team journey is truly incredible!. > Grok-1 Early Access: 3 November 2023.> Grok on X for Premium+ users: 7 December 2023.> Grok-1 Open-Source Release: 17 March 2024.> Grok-1.5 on X: 15 May 2024.> Grok-2 on X: 13 August 2024.> xAI API Launch: 21 October 2024.> Grok-3
30
37
393
Grok-4 and Grok-4 Code benchmarks are š„š„.- 35% on Human last Exam benchmark and 45% with reasoning. - GPQA : 87-88%.- SWE Grok 4 Code : 72-75% N. Note : These are leaked numbers we have to wait for final. h/t : legit_api
šØ Grok-4 prod models detected. Really confused if it's releasing or still delayed.
17
5
103
šØ Grok-4 prod models detected. Really confused if it's releasing or still delayed.
16
15
338