Nikhil Sharma | AI Builder

Building intelligent software with AI. Full-stack AI applications, automation & backend systems. Turning complex ideas into working products.

[Nikhil Sharma | AI Builder @SharNiksX]

I spent my weekend teaching a small Qwen model Bollywood. 😄
Not prompt engineering.
Not an API wrapper.
An actual fine-tuning experiment.

[Nikhil Sharma | AI Builder @SharNiksX]

My weekend experience summed up. 😂

Wanted to fine-tune an LLM.
Ended up debugging VRAM, RAM, GGUF conversion, Colab limits and hardware constraints.

Apparently, becoming an AI engineer comes with a hardware engineering side quest. 😄

[Nikhil Sharma | AI Builder @SharNiksX]

While working with LLMs, I thought fine-tuning would be the hard part.

Gemma 4 E2B on Colab:
15GB VRAM → training worked
LoRA → generated
GGUF conversion → 💥 Out of RAM

Needed ~15–16GB RAM. Free Colab had ~12GB.

Enough VRAM to train. Not enough RAM to finish.

[Nikhil Sharma | AI Builder @SharNiksX]

One thing about learning AI nobody warned me about:

The hardware bill.

Models keep getting bigger. VRAM keeps climbing.

Buy expensive hardware or rent GPUs from the cloud?

How do individual developers keep learning AI without turning it into a GPU-buying competition?

[Andrej Karpathy @karpathy]

We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks:

Writing. Something I've had success with: Ask your LLM to explain something in ASD-STE100, it's a controlled language specification originally developed for aerospace maintenance documentation. LLMs well-versed in this language and it comes with heavy constraints on clean writing style that I often find a lot more readable. Sometimes I've tried to soften it a bit e.g. ask for "80% of the way to ASD-STE100" because the spec is quite stringent. But even better:

Diagrams / images. Instead of writing, ask your LLM to create a diagram. These can be a lot easier to process, parse, and understand. But even better:

Web pages. Ask for output "in HTML" to get a beautiful, interactive webpage. LLMs are getting really good at frontend and can create beautiful experiences, animations, etc. But even better:

Explainer videos. The output format I am most bullish on is fully custom / bespoke explainer videos generated on any arbitrary topic. Experiment with things like "Create a 3b1b style video explainer on X. Use my ElevenLabs API key for audio narration". (you'd need an API key for the latter or you can ask your LLM to find you decent free alternatives that use your local compute). This is actually starting to work!

In summary:
- As LLMs get better, they will do more and more of the legwork autonomously, and a lot more of our work will rise up the abstractions into oversight and understanding.
- Luckily, LLMs can help here too because as intelligence and code are increasingly abundant, you can ask for large, custom, discardable software artifacts (e.g. web apps, video explainers) that would have never made sense to create before. Push the boundaries here and you'll be surprised.

[Nikhil Sharma | AI Builder @SharNiksX]

@Ishwarinfra Yes, needs more attention than we plan for

[Nikhil Sharma | AI Builder @SharNiksX]

AI agents introduce a weird backend failure mode:

200 OK

Tool succeeded.

No exception.

No timeout.

And the agent still did the wrong thing.

Traditional observability asks:
Did the system work?

Agents force us to ask:
Did the decision work?

That’s a very different problem.

[Nikhil Sharma | AI Builder @SharNiksX]

@GoldilocksOrbit Can’t agree more

[Nikhil Sharma | AI Builder @SharNiksX]

What? 1 Million Output tokens 🤯

For context Claude / ChatGPT has around 128k

[Google @Google]

Today we’re introducing Gemini 4 Argon.

It delivers frontier performance in complex workflows across real-world software engineering, knowledge work, and cybersecurity defense with an industry-leading 1M token output limit.

[Nikhil Sharma | AI Builder @SharNiksX]

Will be interesting to see when Gemini 4 Argon is finally released to public

Gemini has finally woken up.
Once September ended. 😂

[Sundar Pichai @sundarpichai]

Lots of discussion out there about our next model(!), so I wanted to give an early look as soon as possible. Introducing Gemini 4 Argon!

It shows frontier performance in complex workflows, cyber defense and software engineering. Teams are using it extensively at Google, from coding to quantum computing, great feedback.

Here’s a look at the benchmarks:

[Nikhil Sharma | AI Builder @SharNiksX]

Asking for a friend. 👀

[Nikhil Sharma | AI Builder @SharNiksX]

Wake up, Gemini. September is almost over. 😴

Anthropic: Claude 5.5
OpenAI: GPT-6
New models. New benchmarks. New leaderboard drama.

Google: 👀

Gemini 4, where are you?

🎵 Wake me up when September ends.

[Nikhil Sharma | AI Builder @SharNiksX]

While reading about Jev, a thought crossed my mind:
Maybe AI systems need to learn when not to think
Not every decision needs an LLM
Some can be handled by rules. Some need reasoning The uncertain ones need a human
Don't spend expensive computation when a simpler mechanism works

[Nikhil Sharma | AI Builder @SharNiksX]

When the OGs says something “pay attention”… 👀

[Andrew Ng @AndrewYNg]

The loudest voices stoking fears about AI dangers have made tremendous headway in the past two weeks. AI technology has not taken some unexpected, dangerous turn, but the hype around it — propelled by what appears to be a well orchestrated PR campaign — has drummed up considerable fear. I worry that it represents a setback for our field.

I have written frequently that fears of AI are overhyped. AI’s capabilities can be uncannily human-like and unpredictable, and it’s rational to worry when people who are directly involved express concerns. But I see the problems as a sign of the engineering work that ahead, rather than insurmountable barriers or the sky falling. AI technology continues to advance — which is a good thing! — but technical advances, poorly understood by the public, give those who seek to generate hype repeated opportunities to do so.

First, I don’t see any step up in the risk of human extinction from AI compared to a few months ago. The theories about this remain the same fantastical, science fiction scenarios as a few months ago. The biggest change in AI risk is its cybersecurity capabilities — a topic which we should take seriously — but this, too, will not lead to the end of the world.

The most notable recent event leading to increased fear was when an OpenAI team deployed an agent swarm that hacked into Hugging Face. Much of the popular press contained significant hype. For example, some publications reported that a swarm of 1,200 agents carried out the attack. While this was technically accurate, as I write this, I have about 1,300 processes running on my laptop. Yes, the ability to get large swarms of agents to work in parallel on a task is a significant technical advance, And, in computing, many processes run at the same time. So this shouldn’t be seen as some magical capability.

Additionally, OpenAI’s buggy sandboxing and monitoring processes were key to enabling this incident. Fixing these bugs and putting in place improved monitoring would be appropriate fixes, not pausing AI. There are many well known ways to attack software systems. The main advantage of AI agents is that they are relentless. They will tirelessly try many tactics — and have the patience to chain vulnerabilities together — that previously would have taken an infeasible amount of human effort. But in the long term, I believe the advantage will lie with defenders (because they have more information with which to identify bugs, which they can fix), but the cyber-threat landscape has changed significantly. There are still bottlenecks to identifying and exploiting a vulnerability. AI agents still have to try a lot of things to see what works, and taking these actions takes time and might be detected by defenders. This is why, even though it is now easy to obtain versions of leading open weight models that have had their guardrails removed or weakened, so they will not refuse to try to execute cyber attacks, the world has not ended.

I am also concerned about the anthropomorphization of AI in a lot of reporting, where LLMs and agents are unnecessarily treated as if they were people. If I wield a hammer, miss a nail, and accidentally dent the wall, it’s not the fault of the hammer. The problem lies in how I used the hammer. Similarly, if I prompt an agent and it hacks into someone else’s system, the responsibility lies with me, not the agent.

Of course, we want to build systems that are as safe and predictable as possible. (For example, an unsafe hammer would be one whose head randomly flies off under normal use.) Today’s agentic systems are not predictable, but I see no reason why, by applying sound engineering practices, we won’t be able to make them extremely safe to use. One new element in the forecasts of AI-enabled doom is AI companies disclaiming responsibility for their own products. “I didn’t do it; my out-of-control agent did!” There’s a balance to be struck between the responsibility of the tool maker and the tool user, but when something goes wrong, let’s hold the people building and/or using the hammer responsible, rather than the hammer. (By the way, if you’re worried about AI bioweapon risk, David Bellamy has a great post on why this, too, is overhyped. Briefly, the bottleneck in building a bioweapon is not intelligence, but lab work and manufacturing.)

Pausing AI progress will create much more harm than benefit. First, our adversaries will certainly not slow down. Second, engineering requires discovering problems empirically so we can fix them. If we pause AI by a decade, we will also delay finding and implementing safety engineering fixes by about the same duration.

Of course, the incentive to stoke fears — for regulatory capture, to garner attention, or to make one’s technology seem more powerful — remains the same as before. Disclaiming responsibility is a new one. Taking a hard technical look at the actual risks however, I see little factual basis for the degree of fear that’s been stoked up. We still have hard research and engineering work ahead to improve AI safety, but the beneficial applications continue to vastly outweigh the risks, and we should keep building.

[Original text (with links): https://t.co/jni2tWazAH ]

[Nikhil Sharma | AI Builder @SharNiksX]

Seems useful 🤔

[Dan McAteer @daniel_mac8]

Claude Coders: run this Opus 5.5 instruction cleanup prompt *NOW*.

'/claude-api prompt-audit'

Optimizes your skills, AGENTS.md, CLAUDE.md and removes anti-patterns that restrict Opus 5.5.

The results are amazing. Best 5 mins you'll spend.

From @RLanceMartin.

[Nikhil Sharma | AI Builder @SharNiksX]

Very Useful content

[Rahul @sairahul1]

Fine-tuning is about to become one of the most valuable AI engineering skills.

Not because everyone needs a custom model.

But because the people who understand how models learn from data will build things others can’t.

Full guide: https://t.co/oIfvJsZgea

[Nikhil Sharma | AI Builder @SharNiksX]

@tusharshekher @pikaso_me screenshot this

[Nikhil Sharma | AI Builder @SharNiksX]

@Swiggy @zomato I think this might be the reason of the cancellation policy where they don’t allow us to cancel, not because of food waste but because they know people might order from somewhere else in meantime in frustration and hunger if they allow cancellation. @deepigoyal

[Nikhil Sharma | AI Builder @SharNiksX]

After further interaction I found they merged my order with someone. Atleast @Swiggy does this with permission. What is the use of gold membership and all with @zomato

[Nikhil Sharma | AI Builder @SharNiksX]

@jammypants4 In one glance it looked like your pic 😂 @jammypants4