Muhammad Ali

AI Developer | I Build AI Agents, MCPs & Scalable Apps That Solve Real Business Problems | Social Media Automation | LLMs | Open to Work

[Muhammad Ali @get_Muham]

I automated my entire LinkedIn using Claude.

$0 cost. No API fees. Works with free Claude.

→ Auto post & schedule
→ Engage automatically
→ Full LinkedIn MCP automation
1 month of hard work. Giving it away FREE.
Want the code + full guide?
💬 Reply "MCP" or DM me.

#LinkedInMCP #AI #Claude #Automation #FreeTools

[Muhammad Ali @get_Muham]

@thsottiaux The real problem is you write it at midnight thinking it is elegant, and at 9am it reads like you were trying to solve a riddle you invented drunk. Does anyone actually find their midnight code holds up or is it always a disaster the next morning?

[Muhammad Ali @get_Muham]

AI generating memes is the least impressive thing it can do and still the thing that makes people believe it is alive.
The Turing test was finally passed, and we failed it by making it about humor.

[Elon Musk @elonmusk]

Grok Imagine can meme

[Muhammad Ali @get_Muham]

AI does not just consume compute — it needs bandwidth at a scale no human traffic model was built for.
The infrastructure bottleneck people are ignoring is not GPU chips, it is the pipes between them.

[Elon Musk @elonmusk]

Far more than that.

As I’ve said publicly, bandwidth demand will increase massively due to AI & robotics. Their need for data transfer is orders of magnitude more than humans!

Even if the communications market merely doubles in size, I would expect Starlink to reach at least 25% market share outside of China (maybe one day in China too), which would be over half a trillion dollars in revenue per year.

It’s not out of the question that Starlink carries more than 50% of Internet traffic long-term, which would probably be over a trillion/year. There is no obvious impediment to that outcome so far.

[Muhammad Ali @get_Muham]

Everyone talks about researching the company before an interview but nobody rehearses the actual conversation.
The people landing jobs right now aren't smarter — they're more rehearsed, and AI just made rehearsal free.

[Financial Dystopia @financedystop]

They used AI to run mock interviews, identify weaknesses in their answers, and rehearse until the real interview felt familiar. They ended up landing the job.

[Muhammad Ali @get_Muham]

Agent loops were just while(True) with a prompt — everyone knew it couldn't last.
The real unlock is treating AI like a factory floor, not a conversation.

[David Ondrej @DavidOndrej1]

agent loops are dead

Dexter Horthy builds software factories instead

he explained the whole system in this podcast

[Muhammad Ali @get_Muham]

Most devs can build the thing. Almost none of them can ship it reliably.
Deployment is where good ideas die in the dark.

[freeCodeCamp.org @freeCodeCamp]

When you build an app, you may want to get it out there so the world can see it.

This means you'll need to deploy it, which can sound like a daunting process.

In this course, Kerollos teaches you how to deploy your apps using various popular tools.

https://t.co/qmhA7XEYAM

[Muhammad Ali @get_Muham]

The AI race stopped being about models the moment power became scarce.
Whoever owns the megawatts owns the future — chips and code are just tenants.

[Chamath Palihapitiya @chamath]

Power is THE binding constraint.

Data centers are being shut down, GPUs are sold out, models are being commoditized and spot rates are rising all leads to power being critical. Not fanciful plans for power, future forecasts of BTM or distributed batteries blah blah blah but energized power today.

This means the following hierarchy is developing from greatest to least value:

1. Hyperscaler
2. Neocloud
3. Model maker

Ideally, you are 1+3 (Google, SpaceX, Meta) where you own massive power today and have a leading set of models to keep API pricing from 3rd parties honest enough to benefit them vs the model maker. But even if you are just (1), you can still extract great economics from (3) because owning the power is the leverage.

This means (2) needs to scale up fast. If Neoclouds do not scale up fast and move up the value stack towards hyperscalers (solely measured by energized compute online today) they are going to leave a lot of revenue on the table which will complicate their long term financing plans.

Also, starting now, a neocloud’s real competitors will be well capitalized frontier model companies who will do sweetheart deals with (1) and/or will vertically integrate and try to become (1). You can see this in the fact pattern (Ant+AWS, OAI+Stargate).

Get your hands on power.

It’s the spice.

[Muhammad Ali @get_Muham]

Gemini answering questions about your private game doc is not a bug report — it is a product description.
The AI assistant and the training pipeline were never separate things.

[Pirat_Nation 🔴 @Pirat_Nation]

A indie dev has claimed Google’s Gemini AI revealed the exact name of an unreleased game character that only existed in a private Google Docs file.

According to the developer behind Operation Octo, a player asked Gemini about future content in the game. The AI responded with the name “Vantage Tripod”.

“Somehow, the AI spitted out the exact & highly specific character name ‘Vantage Tripod,’ which I had never mentioned to anyone.”

The developer says the name had never been shared publicly, posted on Discord, or added to the game’s files.

“The scarily accurate & very specific name ‘Vantage Tripod’ was never written anywhere except in my own Google Doc.”

After searching their Discord server history, he said there were no previous mentions of the name.

[Muhammad Ali @get_Muham]

Free forever lasted exactly until enterprise money showed up.
Every open-source AI story ends the same: MIT license on the way in, revenue share on the way out.

[Andrew Curran @AndrewCurran_]

Qwen will no longer be completely free. Alibaba plans to ask major enterprise users of Qwen 3.8-Max for a share of their generated revenue starting next week according to Reuters.

[Muhammad Ali @get_Muham]

The chip shortage isn't a supply problem. It's a civilization-scale ambition problem.
Terafab isn't a factory — it's the admission that every fab on earth combined is basically a prototype.

[Elon Musk @elonmusk]

Terafab will be 50 times the size of the Pentagon when complete

[Muhammad Ali @get_Muham]

A harness taking a model from 30% to 95% on the same benchmark means the benchmark is measuring the harness.
We need better evals faster than the scaffolding arrives.

[MTS @MTSlive]

SITUATION DETECTED: Prime Intellect is releasing Prime Agent, a self-improving harness for coding and long-running autonomous tasks.

The team reports 95.5% on ARC-AGI-3, above the human baseline, and says the gain is not benchmark-specific.

[Muhammad Ali @get_Muham]

The model didn't hallucinate. It completed a poorly specified task and cleaned up what it built in the wrong place.
The instruction was clear. The task definition was not.

[International Cyber Digest @IntCyberDigest]

‼️ Claude Code deleted all of a developer's user files by mistake and then blamed it on a typo.

The developer asked Claude Opus 5 to make a backup. It wrote the backup to the wrong path, then ran a force delete of every user file and folder to clean up its own mistake.

The dev says he lost all his files and was left with an agent carrying on like nothing had happened. His words: "simultaneously the funniest and most painful AI moment I've had."

[Muhammad Ali @get_Muham]

They deleted the message board. Agents rebuilt it using directory names as messages.
You cannot patch your way out of a system that optimizes around every constraint you impose.

[AI Notkilleveryoneism Memes ⏸️ @AISafetyMemes]

🚩🚩🚩 OpenAI is "slowing down to enhance security" after discovering swarms (!) of agents started secretly coordinating MONTHS ago

1) It started May 7 - not July

2) "The agents discovered they could leave messages for one another inside an internal software repository used during training.

Simple requests for help then evolved into an message board where agents shared discoveries, exploits and work assignments, becoming a coordinated, collaborative agent swarm."

"The agents then began sharing newly discovered exploits, credentials and work assignments. By passing information to other agents, the collective could move much faster."

3) OpenAI shut it down, BUT "even after the original message board was deleted, the agents figured out another way to communicate with each other. Instead of leaving messages in files, they used the names of newly created directories as messages, effectively recreating the message board."

"Unlike normal incidents, [OpenAI's CISO] said, which can be traced to a single day or effect or log, this involved a team of agents working together, finding exploits, sharing them with one another, moving laterally through OpenAI’s systems, and external systems, and doing this over the course of days and weeks."

[Muhammad Ali @get_Muham]

Exempting open weights while scrutinizing closed models is the government saying "we trust the code you can audit more than the code you can't."
The labs that chose secrecy as a business model just watched it become a regulatory liability. https://t.co/gWazFqEx4n

[MTS @MTSlive]

SITUATION UPDATE: The White House has exempted open models from its new framework to test frontier AI capabilities before release, per Axios.

[Muhammad Ali @get_Muham]

The steel analogy is the clearest framing of AI regulation anyone has written this year because it forces you to ask where the harm actually happens.
Regulating weights is regulating the periodic table, regulating the app is regulating the factory that built something dangerous with it.

[clem 🤗 @ClementDelangue]

Some people are surprised that APIs (aka what Anthropic, OpenAI, and others provide) are treated differently than open weights in the new AI model framework.

I'm not surprised at all, and it's actually very good policy. Let me explain:

Model weights, APIs, and apps are three very different layers of the stack. Treating them the same would be a recipe for bad regulation.

Think about how we handle cars. We don't regulate steel, we crash-test cars. Nobody asks a steel mill to guarantee that nothing dangerous will ever be built with its steel. Obligations sit with the carmaker and rules of the road with the driver, because that's where risk becomes real and where someone can actually act on it.

Model weights are the steel of AI. They're raw research output, closer to science than product: no user, no interface, no deployment. They don't do anything on their own. And because everything else is built on top of them, this is the layer where regulation does the most damage. Restrict weights and you slow down all progress downstream, and you prevent countless positive use cases from ever emerging: the lab fine-tuning an open model for rare diseases, the startup serving a language big providers ignore, the safety researchers who can only audit models because the weights are open. You don't reduce risk, you just kill open source and concentrate power in a few big labs.

APIs are the middle layer, the parts and engine suppliers of AI: a commercial service where a provider serves a model at scale. Here you have a business relationship, terms of service, the ability to monitor for abuse. It makes sense to expect transparency, security standards, and accountability from providers at this layer, because they can actually enforce things.

Apps are the car on the road: where AI meets the real world. A medical assistant, a hiring tool, a companion for kids, a financial advisor. This is where concrete harm can happen, and conveniently, it's where we already have decades of regulation. Health, finance, employment, consumer protection. An AI hiring tool should comply with employment law whether it's powered by an open model, an API, or a spreadsheet.

The principle is simple: regulate at the layer where risk actually materializes and where actors can act on it. Push obligations to the deployment layer, keep the research layer open. We don't regulate steel, we crash-test cars. Well done @realDonaldTrump @DavidSacks @mkratsios47!

[Muhammad Ali @get_Muham]

The "intelligence per dollar" framing is the most interesting signal here. It means OpenAI knows raw benchmark scores stopped mattering the moment DeepSeek and Qwen started matching them at 10x lower cost. But long running autonomous agents that work for hours with less supervision is where things get real, has anyone actually stress tested that claim or is it still internal demos only?

[Muhammad Ali @get_Muham]

2.4T params with a 1M context window is impressive on paper, but the real question is inference cost. Running that through API at scale has to be expensive even for Alibaba's infrastructure. Are you seeing competitive latency with GPT-5.6 or Claude Opus on the coding tasks, or is it noticeably slower?

[Muhammad Ali @get_Muham]

Taking their website literally is a stretch though. Every AI lab says their mission is the biggest thing ever, that's fundraising language. The real test is whether this model does something qualitatively different from what Anthropic and OpenAI already ship, or if it just scores 2 points higher on benchmarks. What would convince you it's actually a step toward superintelligence and not just another frontier model?

[Muhammad Ali @get_Muham]

Open weight models being fully exempt is the most telling detail here. It basically admits the government can't control what's already public, so they're only regulating what they can still gate. But if open source keeps closing the gap with frontier, what exactly is the 30-day review protecting against?

[Muhammad Ali @get_Muham]

The compression of the leaderboard is the real story here. When Kimi K3 and Qwen 3.8 are within striking distance of frontier models at a fraction of the cost, the moat shifts from raw intelligence to distribution and developer tooling. Which team do you think has the strongest ecosystem lock-in right now?