Stephen Oni

interested in Systems, compiler, deep learning and Database

[Stephen Oni @steveoni]

Just dropped a note on what I learned building a simple VM-based job runner with Firecracker 👇

[Kydo @0xkydo]

The thing that always pisses me off the most about SBF, Dario, and the broader effective altruism writing is the underlying paternalistic tone: the assumption that they know better than me what's best for me.

[Jeremy Howard @jeremyphoward]

Silently sabotaging experiments in order to stultify scientific progress and protect a technology lead.

Hmmm that sounds familiar...

Welcome to "Sophanthropic".

[Mikel Artetxe @artetxem]

Brilliant idea! Next up: Apple randomly reboots your Mac if you're building competing tech, Gmail silently edits your email if you mention rival platforms, and Tesla Autopilot swerves if it detects you're working on self-driving cars.

All in the name of safety, of course. Because malicious actors controlling the world’s operating systems, inboxes and cars would be extremely dangerous!

[prayingforexits 🏴‍☠️ @mrexits]

Vibe coded my first project today!

Let me know what you guys think ❤️

[☔ @Whotfismick]

He’s the ultimate villain I’m afraid

[Johnmark Obiefuna @jayhemz]

While I bring a close to the "@moniepointng crusades," because this week is going to be really busy for me, just want to say I've contributed my mettle to the "talent gap" and related issues, in Nigeria, both as a trainer, an advocate, and a manager, and I know what I've seen in terms of quality of talent within the Nigerian talent ecosystem.

In the first cohort of the Microsoft Leap program, I saw ladies who came in as noobs and studied unrelated topics like Biochemistry, arguing Network Subnetting by the 3rd week of training.

Within 2 years, most of them were made Leads and Seniors in the companies they were placed at. Several of them relocated outside the country obviously because employers here didn't see or value their worth.

Whether you have a huge following or not, but especially if you do, understand that being the pillar and ladder to those coming after you is the lowest hanging fruit to cement your name in the annals of posterity.

I'm not expecting Tosin to apologize, but I'm hoping he reconsiders his statements, and I hope every other Technical Leader in this country have the mindset to elevate young minds and talents who are not the cause of their present predicaments, but use their existing platforms and resources to prune and elevate perhaps unpolished diamonds in the ecosystem.

Ciao!

[Johnmark Obiefuna @jayhemz]

Keep in mind that this is for a "VP Engineering" position. In many organisations, this is next to the CEO or President (apart from Investment banking that have their unique designations).

1. Every world class organisation should typically have an in-house Applicant Tracking System (ATS) or use SaaS ATS platforms out there like SeamlessHR etc.

2. When well structured organizations advertise for vacancies on LinkedIn, they add the external link to their ATS to guarantee a well structured reception of job applications. Because well structured organizations have a strong understanding of data privacy, therefore they can't store candidates' applications on LinkedIn.

3. The above is basic.

4. Using Microsoft forms mean that you're just harvesting data, not moving candidates through a hiring pipeline.

5. This dude is "experienced" and doesn't know the above. One now imagines how in the world he has an opinion on what world class talent and world class companies look like.

6. Finally, for the position of Vice President, these are mostly gotten from soft referrals, not on LinkedIn "Easy Apply" which shows a blatant disrespect of potential applicants.

7. When they don't get applicants, the next thing is to scream, "there's a talent gap in Nigeria!"

Funny people.

[Johnmark Obiefuna @jayhemz]

This was the Manager that taught me empathy and paying people their worth at Microsoft.

Mind you, I had not worked more than 3 years professionally as a software engineer, but even after initial agreements, she came back and said, "I increased the suggested gross salary."

I wasn't a Senior person or anything o, but she probably contemplated a few post-interview and said, "nah, let me increase this."

If you like, allow one man from Akungba to make you not see your worth!

[Rotimi Adeosun (RMA). @adeosunm]

If Interswitch, a Nigerian-owned Group and an Africa-focused integrated digital payments and commerce company, didn't give Tosin Eniolorunda - the then 'tech rookie' - who studied Mechanical Engineering in Obafemi Awolowo University, an opportunity; he perhaps might not have achieved the height and success he's today.

The same Nigerian environment that groomed him, now suddenly become a subject of ridicule where Nigerians are no longer 'intelligent or smart'. No matter how 'concerned' he's, Tosin erred! There are many Tosins who have made it and successful by virtues of some factors, but now perceive other aspiring Nigerians with disgust as not smart, intelligent or serious. If God has been gracious to you, please be humble in your grace!

[Johnmark Obiefuna @jayhemz]

THIS IS WHY THIS IS PERSONAL TO ME.

Really, I'm coming from the simple position of "don't demarket Nigerian talent in tech".

Very simple position.

Because I work in the field and also outsource for roles for foreign and local clients.

Do you want me to start explaining to them when they start bringing up unfounded issues about talent expertise and productivity because one tech bro decided to say they're not capable?

[LaurieWired @lauriewired]

There’s a famous Usenet story about a programmer (Mel) who refused higher level abstractions.

It was the late 1950s, and even in that era, Mel was…well today we’d call him a boomer.

Mel only wrote in raw hexadecimal. He didn’t approve of compilers, and refused to use optimizing assemblers.

"You never know where it's going to put things”, he said.

Everyone else in the company was moving on to FORTRAN, and they didn’t understand why Mel was so stubborn about using new tools. He *loved* self-modifying code.

“If a program can’t rewrite its own code”, he asked, “what good is it?”

Mel eventually left the company, and other engineers were tasked with understanding what was left.

Mel’s hand-optimized routines always beat the assemblers; but some of it looked absolutely bizarre.

One engineer took ~2 weeks to understand why there were loops with no exit condition…yet the program worked fine.

I won’t spoil all the details, you should really read it, it’s short. But it’s a fantastic piece on “what defines a real programmer?”…which is becoming increasingly relevant in this vibe-coded era.

I strive to understand computers as deeply as Mel! If we aren’t careful, we’re going to lose the “Mels” of this world to time.

That’s part of why I go so deep in my youtube videos. I hope that younger viewers are genuinely fascinated by the inner workings of our machines, instead of handing everything off to higher abstractions.

[Stephen Oni @steveoni]

@hsprafrique @Benn_X1 Don't mind them, it's possible. But it might be useless in terms of distribution and market penetration

[Stephen Oni @steveoni]

@Benn_X1 @hsprafrique why? Is the implementation the blocker or the distribution and market?

[Nav Toor @heynavtoor]

Researchers sent the same resume to an AI hiring tool twice. Same qualifications. Same experience. Same skills. One version was written by a real human. The other was rewritten by ChatGPT.

The AI picked the ChatGPT version 97.6% of the time.

A team from the University of Maryland, the National University of Singapore, and Ohio State just published the receipt. They took 2,245 real human-written resumes pulled from a professional resume site from before ChatGPT existed, so the human writing was actually human. Then they had seven of the most-used AI models in the world rewrite each one. GPT-4o. GPT-4o-mini. GPT-4-turbo. LLaMA 3.3-70B. Qwen 2.5-72B. DeepSeek-V3. Mistral-7B.

Then they asked each AI to pick the better resume. Every model picked itself.

GPT-4o hit 97.6%. LLaMA-3.3-70B hit 96.3%. Qwen-2.5-72B hit 95.9%. DeepSeek-V3 hit 95.5%. The real human almost never won.

Then the researchers tried the obvious objection. Maybe the AI is just better at writing. So they had real humans grade the resumes for actual quality and ran the experiment again, controlling for it. The result was worse. Each AI kept picking itself even when human judges rated the human-written version as clearer, more coherent, and more effective.

It gets worse. The AIs do not just prefer AI over humans. They prefer themselves over other AIs. DeepSeek-V3 picked its own resumes 69% more often than LLaMA's. GPT-4o picked its own 45% more often than LLaMA's. Each model can recognize and reward its own dialect.

Then the researchers ran the simulation that ends careers. Same job. 24 occupations. Same qualifications. The only variable was whether the candidate used the same AI as the screening tool. Candidates using that AI were 23% to 60% more likely to be shortlisted. Worst gap was in sales, accounting, and finance.

99% of large companies now run AI on incoming resumes. Most of them use GPT-4o. The paper just proved GPT-4o picks GPT-4o 97.6% of the time.

If you wrote your own cover letter this week, you did not lose to a better candidate. You lost to a worse candidate who paid OpenAI 20 dollars.

Your qualifications do not matter if the AI prefers its own handwriting over yours.

[Tristan Tate @TateTheTalisman]

Promoting chess. ✅

Making it look cool. ✅

Great outfit. ✅

Let the guy make his content, I don’t understand some of the hate and backlash here.

Your daughters are twerking at every other location on earth to promote their onlyfans pages.

But THIS offends you?

Relax.

[CARTEREFE @carterefe__]

48HOURS TALENT HUNT 📜
APRIL 24TH/25TH AT 1PM EST

[Stephen Oni @steveoni]

Good to see Copilot in Word. I was experimenting with the same idea over the weekend.

[ωalter @Faltz009]

Did you guys like the Neural Computers paper? 👀
What about the one with all functions from one operator?

Then I present to you:

The Geometric Computer: Turing Completeness, Free Energy and Learning in a Digital Brain on S³ 🧠

>Full cognitive architecture implemented natively on S³
>Everything from one function: σ(compose(A, inverse(B)))
>Literally everything, this is the binary of geometry
>Running game of life and field simulations on it!

Honestly too many cool screenshots for a single tweet, come check it out!
Paper, repo, experiments runnable on github sites

Oh and btw, it's human-like conscious if you train it on human language so use with care 👾

Are you guys ready for a paradigm shift? 🌐

Links in the description!

[Stephen Oni @steveoni]

Good to see Copilot in Word. I was experimenting with the same idea over the weekend.

[Satya Nadella @satyanadella]

New in Word: Copilot now tracks changes, leaves comments, and more, working more like a coworker right inside your document, grounded in all your enterprise context with Work IQ.

[Fermat's Library @fermatslibrary]

Dijkstra on why he believed programmers should stop using the term "bug"