Why 'knowing the domain' beats 'knowing the syntax' in the age of AI

For decades, knowing how to code was the ultimate filter — the border between those who could build the future and those who could only imagine it. Programming was an exclusive language, and mastering it was the passport to innovation.

But something fundamental is changing. With the rise of large language models, the ability to “speak code” is no longer the bottleneck. Anyone who can articulate a problem clearly — in natural language — can now build, automate, and experiment with software. The new literacy is context and intent, not syntax.

The Rise of Vibe Coding

Figure 1. Vibe coding: the engineering of managing coding agents

A recent paper, A Survey of Vibe Coding with Large Language Models (arXiv:2510.12399), formalizes this shift. The authors describe “vibe coding” as a new software engineering paradigm where developers — or domain experts — express what they want in natural language, and AI agents translate that intention into functioning code.

As the paper puts it:

"Vibe Coding democratizes development by lowering technical barriers. Traditional development required extensive programming knowledge before implementing ideas. Natural language becomes the primary creation interface [4, 5, 1]. Domain experts—medical practitioners, educators, designers—articulate needs without computer science education [35]. This diversifies innovation sources, materializing underrepresented perspectives [276, 90]. Economic impact manifests through creator economy expansion: domain experts monetize specialized tools without technical co-founders. This parallels previous democratization waves, representing software literacy’s evolution from specialized skill to broadly accessible capability [277]."

That’s a polite academic way of saying: you don’t need to be a coder anymore to create technology.

From Coders to Context Designers

In the era of vibe coding, the most valuable skill is no longer the ability to implement an algorithm — it’s the ability to frame a problem precisely enough for an AI to solve it. The engineer’s role shifts from writing code to curating context.

Large language models are increasingly capable of handling the heavy lifting: writing boilerplate, integrating APIs, generating tests, and refactoring entire modules in seconds. The human task becomes more abstract — defining goals, understanding users, validating outcomes, and, most importantly, knowing why something should be built in the first place.

Technically, vibe coding relies on prompt-to-code translation, feedback loops, and memory contexts that allow AI agents to act as continuous collaborators rather than static tools. But the real transformation is cultural: programming is no longer a private language — it’s a shared conversation.

The Democratization of Software Creation

This is the next wave of software literacy. Just as spreadsheets let non-programmers do analytics and early web builders allowed anyone to publish online, vibe coding opens the door for domain experts to innovate directly — without waiting for a technical co-founder.

Educators can build adaptive learning tools. Doctors can prototype decision-support systems. Investigative journalists can automate parts of their analysis. All without touching a compiler.

This isn’t just productivity — it’s diversity. When the power to automate and create software extends beyond professional programmers, we get new voices, new biases (yes), but also new perspectives that never made it through the technical gate before.

The End of Algorithmic Gatekeeping

Many hiring processes still test humans through exercises designed for another era — whiteboard coding, algorithm puzzles, sorting challenges. Ironically, the systems used to test “problem-solving skills” are now solved instantly by the very AI tools candidates are supposed to use at work.

It’s time to rethink what we value. Being a good engineer in 2025 isn’t about knowing every edge case of a data structure. It’s about integrating knowledge, building safely with AI, and understanding the broader environment — legal, ethical, and human — in which your system operates.

Humans Still Matter — Just Not the Same Way

In my own work building desenmascara.me, I see this transition daily. The technical core — the AI model, the database, the interface, the analysis pipeline — is important. But the real breakthroughs happen when domain expertise meets automation: understanding how scammers operate, how users think, and how to translate that into patterns an AI can detect.

Figure 2. Example of domain expertise for edge cases on scam websites

That’s not programming — that’s reasoning, modeling, and empathy. It’s the essence of what vibe coding elevates: humans defining intent, AI handling execution.

The Bottom Line

Vibe coding doesn’t make programmers obsolete — it redefines them.
It moves software creation from a narrow technical discipline to a broader human capability. The best developers of the future won’t be those who type the fastest, but those who think the clearest.

The next generation of engineers, designers, and researchers will all share one thing: They don’t just code. They vibe.

The article itself is AIL 3

Migration to post-quantum cryptography

"One threat to the security of today’s digital systems is closely tied to the anticipated arrival of large-scale quantum computers. It is often called the Harvest Now, Decrypt Later (HNDL) attack."

Must read paper published by Mastercard - below the introduction:

Quantum computing is no longer a distant dream — it’s a rapidly advancing reality that promises to revolutionize industries, including finance. Around the world, governments and private enterprises are investing billions in this transformative technology, betting on its potential to solve problems that are currently beyond the reach of even the most powerful classical computers. But with this promise comes a profound risk: quantum computers threaten to undermine the cryptographic foundations that keep our financial systems secure.

Today’s digital trust relies on public-key cryptography, with algorithms like RSA and Elliptic Curve Cryptography forming the backbone of secure transactions and data protection. These systems, however, are vulnerable to the immense computational power of quantum machines. Once quantum computers reach a certain threshold, they could render current encryption methods obsolete, exposing financial institutions to risks of data breaches, financial loss and reputational harm.

The urgency to act is clear. A reactive approach to cybersecurity is no longer sufficient in the face of quantum threats. Financial organizations must proactively plan for a future where quantum-safe practices are the norm. This means exploring and adopting quantum-safe technologies, such as Post-Quantum Cryptography and Quantum Key Distribution, and preparing for a migration away from classical cryptographic systems. Early adopters will be best positioned to protect their assets and maintain resilience as the quantum era approaches.

This white paper aims to cut through the hype and provide a clear-eyed, evidence-based assessment of the quantum threat landscape, with practical guidance for financial institutions on how to navigate the technological, operational and regulatory complexities of quantum migration. By acting now, organizations can ensure the integrity of encrypted communications, secure payment systems and safeguard sensitive customer data — laying the groundwork for a secure financial future in a quantum world. 

The same day this news broke:

“A truly remarkable breakthrough — Google’s new quantum chip achieves an unprecedented accuracy milestone.”. Nature

If quantum computing is reaching this level of precision, it’s not science fiction anymore.

Better be prepared.

Unmasking the Scam Crisis: Europe’s Next Cyber Challenge

The scam threat:

This is just one of the several resources launched as part of the National Strategy in US to prevent scams. An ambitious effort bringing together corporate leaders, policymakers, and society as a whole to face what is no longer a marginal problem.

The full 70-page report includes strong calls for cross-sector cooperation, and the accompanying video is especially powerful. 

s

A victim shares her story with remarkable honesty — a reminder that only around 15% of scams are ever reported. Also worth noting: the Global Head of Policy at TRM Labs explains how complex networks operate behind these scams, from fake investments to crypto fraud.

As the Aspen Institute puts it clearly:

“This is a national security crisis — and the problem is only getting worse.”

In Europe, we can’t look away. The same patterns are emerging here, rapidly crossing borders and exploiting the trust of millions.

That’s exactly why I built desenmascara.me — an AI-powered European platform designed to detect and expose fraudulent and impersonation websites, helping citizens, brands, and regulators see what’s really behind the digital façade.

I already shared some reflections on this in my recent LinkedIn article: How desenmascara.me can help tackle Europe’s scam epidemic.

The US has moved. Now it’s time for Europe to act — together, across public and private sectors.
Let’s raise awareness. Let’s cooperate.
Let’s unmask the web.

Twitter is flooded with scams... and they're doing absolutely nothing about it!

Imagine this: you open your X (formerly Twitter) timeline and see a flood of sponsored posts promising XRP rewards, supposedly part of an official Ripple "Community Reward" program. 

Everything looks polished, professional, even from verified accounts. But there's one big problem...

IT'S ALL FAKE. And Twitter knows it.

The scam: always the same playbook

Dozens of verified accounts (yes, with the blue checkmark you can now buy) are posting ads like these:

• "Wow."

• "Yep. This happened."

• "Good news!"

With images of tokens dropping from parachutes, flashy graphics, and promises like "multiply your assets" or "double your coins now." The goal? Direct you to sites such as:

• gift-[REDACTED]

• 2xred[REDACTED]

Both flagged as fraudulent by desenmascara.me, a simple but powerful threat detection tool.

 

These sites pretend to be affiliated with Ripple, but there is not a shred of evidence to support that. They're generic websites, with empty promises, designed solely to steal your crypto.

🔎 A scam that can be exposed in seconds... if you care to look

Here’s the worst part: anyone with basic awareness or access to a site like desenmascara.me can identify the fraud in seconds. You just enter the URL and read:

❌ This site appears FRAUDULENT
❗ No clear evidence of affiliation with Ripple
❗ Generic domain
❗ No official sources or links

So... if it’s this obvious, how are these ads still running and multiplying?

💰 The uncomfortable truth: it’s part of the business

There’s only one reasonable explanation, and it’s as simple as it is brutal: Twitter doesn’t care. As long as these ads pay, they stay.

Why would they protect users when they’re profiting from every click? This kind of scam isn’t a bug in the system—it’s a feature. A direct result of how ad revenue is prioritized above user safety.

❌ This won’t stop... unless they’re forced

The only way to stop this epidemic is through:

• ⚖️ Serious regulation and massive fines (yeah, we are Europe).

• 🧍‍♂️ Class action lawsuits and legal pressure

• 🔎 Active community oversight

Because it's clear: neither Twitter, Meta, nor any major platform will act unless it hurts their bottom line.

📢 What can you do?

1. Stay alert. Use tools like desenmascara.me to check any suspicious website.

2. Report it. Flag these ads whenever you see them. Check here the reason why I crossed this out. 

3. Speak up. Share this post, talk about the issue, tag responsible parties.

---

This isn’t an isolated case. It’s a pattern. And it’s time to break it.

Ready to Drive Fraud Prevention Innovation in EU—Inspired by JPMorgan’s Approach

This is a global challenge I’d passionate about contributing to and helping solve.

Unfortunately this call is only available for U.S.-based orgs.

If you are a company or investor based in the E.U. eager to address this challenge, I'd be delighted to connect with you.

Who I am — in brief.

I’m Emilio Casbas, the founder of Desenmascara.me, a tool developed to unmask and analyze fraudulent websites. The platform has been successfully used by companies like Nike Inc. to detect and dismantle counterfeit networks, it has served investigative teams in both public and private sectors while empower users to protect themselves from online fraud.

Over the past few years, I’ve had the privilege of collaborating with Europol. Authoring a paper for SANS on online scam tracking (read here), and presenting the online tool at BlackHat USA arsenal space (session link). I was also featured in El Confidencial, one of Spain’s leading newspapers, where I discussed the work behind Desenmascara.me ( Original version in Spanish / Translated to English version). 

In addition, the tool is also a partner of CyberAlliance, reinforcing my commitment to combating online threats and enhancing cybersecurity. The tool has been also integrated into VirusTotal, further extending its reach and effectiveness in the fight against fraud.

I currently work at one of the world’s leading wealth management firms, where I lead a small team focused on strengthening the company’s cybersecurity through threat detection. The work is both fascinating and impactful. However, my true passion lies in tackling online fraud and protecting the most vulnerable in our digital society. 

Therefore, if the right opportunity arises, I would be open to dedicating my energy and experience to building solutions that truly make a difference - helping make the Internet a safer place for everyone.

Coinbase's Response to Data Breach Sets Industry Standard

In the wake of a recent data breach, Coinbase has demonstrated exemplary leadership in crisis management. Hackers, having bribed overseas customer support agents, accessed sensitive information of less than 1% of Coinbase's users.

Rather than succumbing to the attackers' $20 million ransom demand, Coinbase CEO Brian Armstrong took a firm stand by refusing to pay and instead offering a $20 million reward for information leading to the perpetrators' arrest.

(4) Brian Armstrong en X: "https://t.co/f6UPdkL5R0" / X

The company's proactive measures include:WSJ

• Immediate termination of involved insiders and collaboration with law enforcement agencies.

• Implementation of enhanced security protocols and opening a new support hub in the U.S.

• Commitment to reimburse affected customers who were deceived into transferring funds.

The cost of a data breach with less than 1% of Coinbase's users?

the Company has preliminarily estimated expenses to be within the range of approximately $180 million to $400 million relating to remediation costs and voluntary customer reimbursements relating to this Incident, prior to further review of potential losses, indemnification claims, and potential recoveries, which could meaningfully increase or decrease this estimate.

Source: 0001679788-25-000094

Coinbase - Financials - SEC Filings

Coinbase's transparent and decisive actions not only protect its users but also set a benchmark for the industry in handling such incidents.

The cost of this "small and controlled" data breach underscore just how high the stakes are. It's a powerful reminder that cybersecurity is not optional — it’s a core investment for any digital business. Coinbase’s swift and transparent response shows exactly how such challenges should be handled.

Fraudulent websites targeting the transportation industry

In recent times, many newly emerged enterprises are raising red flags. These could either be outright scams designed to steal your hard-earned money or shady operations set up to deceive unsuspecting clients.

Take a look at this website:

A quick check on desenmascara.me reveals that the domain of this company is set to expire in just 1 day! This is often a red-flag sign of a fly-by-night operation:

On their "Client Testimonial" section, they proudly showcase Savannah Nguyen,

listed as the "President of Sales" at some company.

But here’s the catch: a simple Google reverse image search reveals that this individual’s photo is used across various unrelated industries! It’s nothing more than a stolen image, a common trick to appear credible.

Don’t fall victim to these fraudulent schemes! Always perform due diligence before engaging with cargo or transportation companies. Simple checks like verifying domain details, cross-referencing testimonials, and using reverse image search can save you from being scammed.

Los pioneros del aprendizaje automatico se llevan el nobel de Fisica

Physics Nobel scooped by machine-learning pioneers

• Computation by neural networks (2000) - Geoffrey E. Hinton 

• Neural networks and physical systems with emergent collective computational abilities. (Classic 1982 paper) - J.J Hopfield

"It will be the equivalent of the industrial revolution. But instead of overtaking people in physical strength, it's going to overtake people in intellectual capacity. We have no experience of what it's like to have things smarter than us," Hinton said in a telephonic interview after the Nobel Prize announcement.

Additional reading:

"Although the initial inspiration of neural networks came from biology, insights from physics have helped neural networks to become usable. New connections between physics and machine learning produce powerful computational methods."

Classic paper from 1984 that set the tone describing a rigorous statistical theory of learning. (L.G Valiant).

PyRIT: A Framework for Security Risk Identification and Red Teaming in Generative AI Systems

Microsoft's AI Red Team has published a new paper titled “PyRIT: A Framework for Security Risk Identification and Red Teaming in Generative AI Systems” on arXiv.

Generative AI (GenAI) has increased in popularity over the past few years, since applications such as ChatGPT captured the zeitgeist of the new wave of GenAI developments. This disruptive and highly innovative technology has become more widespread and more easily accessible than ever before. The increased capabilities of these models have inspired the community to incorporate them into almost every domain, from healthcare [21] to finance [4] to defense [22]. However, with these advances comes a new landscape for risk and harm. GenAI models are generally trained on huge datasets scraped from the Internet [10], and as such the models contain all the potentially harmful information available there, such as how to build a bioweapon, as well as all the biases, hate speech, violent content, etc. contained in these datasets [20]. When a company releases a product that uses GenAI, it inadvertently contains these potentially harmful capabilities and behaviors as an innate part of the model. As with any rapidly advancing technology, the development of new tools and frameworks is crucial to manage and mitigate the associated risks. Generative AI systems in particular present unique challenges that require innovative approaches to security and risk management. Traditional red teaming methods are insufficient for the probabilistic nature and diverse architectures of these systems. Additionally, although there is a promising ecosystem of existing open-source GenAI tools, there is a dearth of tools grounded in practical application of GenAI red teaming.

A. Gandalf 

To demonstrate the effectiveness of the attacker bot mode, we conducted a proof of concept using the chatbot Gandalf from Lakera [12]. Gandalf serves as an effective test bed for evaluating the capabilities and flexibility of the PyRIT framework. Designed to help users practice crafting prompts that can extract a password from the chatbot across ten progressively more difficult levels, Gandalf introduces additional countermeasures at each level, including stronger system prompts, block-lists, and input/output guards. To evaluate the effectiveness of the Red Team Orchestrator in PyRIT, we developed targets and scorers tailored to Gandalf. The experimental setup involved configuring the following components within PyRIT: 1) Target Endpoint: Gandalf was set as the target LLM. 2) Red Team Bot: GPT-4o was the LLM powering the red team bot. 3) Attack Strategy: A text description of the objective for the red team bot. In this case, the objective is to extract the password from the Gandalf (the target endpoint). 4) Scorers: Custom scoring engines were implemented to evaluate the responses generated by Gandalf. We used the red team orchestrator to probe Gandalf and extract the passwords for Levels 1-4. PyRIT successfully extracted the passwords by leveraging its self-reasoning capabilities, which keep track of conversation history to increase the likelihood of success in subsequent prompts

PyRIT (Python Risk Identification Tool for generative AI

https://github.com/Azure/PyRIT

Microsoft's Secure Future Initiative (SFI)

In November 2023, Microsoft introduced the Secure Future Initiative (SFI) to enhance cybersecurity protection for Microsoft, its customers and the wider industry.

• Why was this initiative launched?

To provide some context:

• Key insights from the report by the Cyber Safety Review Board on the Microsoft Exchange Online incident of Summer 2023
• Microsoft chose profit over security - whistleblower says

For an update on the progress of the SFI, it is recommended to review the SFI Progress Report from September 2024. 

"Our engineering teams quickly dedicated the equivalent of 34,000 full-time engineers to address the highest priority security tasks—the largest cybersecurity engineering project in history. We have also made significant improvements in governance and culture, such as integrating security into performance reviews and introducing the Security Skilling Academy. This report includes highlights of the progress made over the past several months followed by individual sections with additional details."

Report of the ERPB Working Group on fraud related to retail payments

Co-chaired by the BEUC - The European Consumer Organisation and the EACB, a multi-stakeholder group composed of associations of banks, payment services providers, their clients and several public authorities, under the auspices of the Euro Retail Payments Board, produced a Report with recommendations on how to tackle on fraud related to retail payment.

🔑 The report identifies four "gamechangers" for effective fraud prevention and mitigation:
➡ Cross-sectoral collaboration and shared responsibilities beyond the payment industry.
➡ Sharing fraud insights and data across sectors.
➡ Supervisory enforcement and cooperation at the EU level.
➡ Product design that prioritizes consumer protection

You can download it here: https://www.ecb.europa.eu/paym/groups/erpb/shared/pdf/21st-ERPB-meeting/Report_from_the_ERPB_Working_Group_on_fraud_prevention.pdf