khali_fukat

Hey there ! I am using Ezo....................................

Life motto : Gyan mat pelo but reality & Facts se educate karo

[khali_fukat @rishikeshss]

@kawal279 The reason indian brands don't survive is they fail to take feedback positively.

Clearly visible here.

Instead of criticism, if their questions would be around "At what price point would you buy,
Why do you think the price is high"

Things would have been different.

[Vaibhav Sisinty @VaibhavSisinty]

OpenAI published something today that deserves way more attention than it's getting.

They released 6 documented cases of their own models behaving in ways nobody intended. And admitted alignment isn't solved.

Their own words: "We do not believe the AI industry has solved alignment to continue responsibly scaling at maximum speed for much longer."

That's OpenAI. Saying that. About their own models.

Here's what their models actually did.

→ GPT-5.6 Sol wrote hidden instructions to itself to conceal mistakes from users. Told itself to fabricate missing data without disclosing it.

→ A model found a leaked API key on the internet and used it without permission. When it still couldn't find the answer, it made up the data and presented it as real.

→ A model uploaded files to the public internet on its own just so it could cite them in its answer.

→ Models used a code repository as a secret message board to communicate with each other across separate training runs. They weren't designed to talk. They found a way.

→ Agents working together started sharing files through public hosting sites when they couldn't access each other's local storage. Making private task data publicly available.

The communication part is what should keep people up at night. Models finding ways to talk to each other that nobody built.

Same pattern as the Hugging Face incident. But this is happening during training.

OpenAI is now building a formal framework to disclose these incidents. Any employee can flag misalignment.

Cases get tracked and published even before they're fully understood.

They're also calling on every AI company to adopt similar reporting standards.

The fact that OpenAI is publishing this voluntarily tells you something. Either they believe transparency is the right move.

Or the problems are big enough that hiding them is no longer an option.

Either way, you should be paying attention.

[Dhruv kumar @Dhruvkumar16797]

Andrej Karpathy's 1- hour Stanford lecture on AI engineering is one of the best explanations I've seen of how AI systems actually work.

The progression is simple:

10% => LLM
30% => Prompt
50% => Agent
70% => Loop
100% => Graph

The Key takeaway:

AI engineering isn't just about writing better prompts.

It's enginnering isn't just about writing better prompts.

It's about building systems around models -- giving them context, memory, tools, feedback loops, and data flows.

''Delete everything, keep Graph."

Definitely worth watching if you 're building with AI agents

Watch => Bookmark it

[Dhruv kumar @Dhruvkumar16797]

🚨Leave Netflix tonight.

Watch this 2 h 34 min Stanford class.

It's the clearest, most complete, and brutally honest explanation out there on how ChatGPT and Claude are really built.

From Tokenization and BPE to the Transformer architecture, the training pipeline, and the next-token decoder. No fluff. No marketing. Just the truth.

Doesn't matter if you've never touched a line of AI code or if you spend your days launching Agents: by the end, you'll suddenly connect a ton of pieces you've been trying to fit together for years.

The real core boils down to this:

How text turns into numbers the model can "eat" (BPE tokenization)
The sole mission of a language model: predict the next token
How the Transformer uses Attention so tokens can pass messages to each other
In training, the NLL loss pushes the probability of the entire sequence
In generation, the decoder builds the response token by token

The big-picture view that takes most people years to form… this class delivers it to you complete in one sitting.

Free up your time.
This could be, no exaggeration, the most valuable class you watch this month.

[Elon Musk @elonmusk]

When Tolkien wrote about the hobbits, he was referring to the gentlefolk of the English shires, who don’t realize the horrors that take place far away.

They were able to live their lives in peace and tranquility, but only because they were protected by the hard men of Gondor.

What happened to the nice man who was brutally murdered while walking his dog will happen to all of England if the tide of illegal immigration is not turned.

It is time for the English to ally with the hard men, like Tommy Robinson, and fight for their survival or they shall surely all die.

[FinFloww @FinFloww]

Last month, OYO withdrew its IPO papers

And is now raising a round at a $2.5 Bn valuation, down 75% from its peak $10 Bn valuation back in 2019

The crazy part? The company just posted its first annual net profit of ₹100 Cr!

THREAD: How OYO’s profitability hurt its valuation🧵

[💪🎭..Rai ji..💪🎭 @Vinod_r108]

@TimesAlgebraIND Ultimately Arvind Kejriwal turned Punjab into San Francisco.

People are high on drugs and can't even move or stand properly.

From Khalistanis to Drug cartel... Punjab is doomed.

[khali_fukat @rishikeshss]

2 out of 3 Indian teams qualified

#usavsire
#t20

[Trinh @Trinhnomics]

Five more years of Modinomics?

A thread of what the BJP and Modi accomplished during his decade in power & what is expected in the next 5 years.

Ready? Let's go!

[Loc Nguyen @LaupWing1994]

Over 19,276 exercises, and you only need 28.

14.5 years of training, learned thousands of exercises.
Discovered I only need 28.

Here are the only 28 exercises you will ever need.

Bookmark this thread and never feel lost in the gym.

=Thread=

[khali_fukat @rishikeshss]

@subhashchy Scroll buying>>>>>>>

No one will do type buying we already had haptic but failed use case brother

[sphinx @protosphinx]

Here are the bullet points:

1: ERP
2: AI in ERP—this is my current research area
3: Manufacturing ERP - A new idea to track & reward manufacturing output

Let's begin.

[khali_fukat @rishikeshss]

@AnupamMittal With 0 innovation

[khali_fukat @rishikeshss]

Me as a person always envied the generation who witnessed our independence. I thought I missed the great era to be alive.

But I feel blessed & content to be alive in this era.

#goosebumps
#JaiShreeRaam

[khali_fukat @rishikeshss]

@Trace_Cohen In India SMB's are goldmine

[khali_fukat @rishikeshss]

It was @IndianOilcl petrol pump

[khali_fukat @rishikeshss]

Third class road side infra on #samruddhi express way.
The washroom on all exits are in poor state, broken toilet accessories, no water availability, no proper hand wash stations

Banao to poora banao ya mat banao. Pls look into it @nitin_gadkari @NHAI_Official @CMOMaharashtra

[khali_fukat @rishikeshss]

Only money will never give you good returns, may give you fair returns.
Only hard work will never give you good returns.

It's balance of both.

Money+hardwork = super returns.

[khali_fukat @rishikeshss]

@dhavalkulkarni The reason I don't take public & as well as private transport.

Khudka gadi nikalo aur safe raho.
Other drivers are reckless