Why Current AI models Aren't So Smart [Brave New World]

BBC

Grok is part-smart, part-dumb (so is Elon Musk) 

If you have time to read the article, this is an interesting insight into the limitations of current AI models, and what the next step - or a different way of doing things - could be. It puts current debates on AI in perspective. 

AI-powered humanoid robots are not very good at doing fiddly things that a human being, once properly trained, can do easily. This would include simple tasks such as filling up the dishwasher. Trades such as plumbing are not going to be replaced by machines any time soon. 

The Large Language Models (LLM), on the other hand, are very good at collecting information, analyzing it, and summarizing it: this is what many white-collar jobs entail in sectors such as banking, insurance, logistics, administration, etc. So, at this point, this new industrial revolution is going to hit white collars frontally, when at least some trained blue collars will be spared, paradoxically. 

At the end of the day, even sophisticated AI machines are just intelligent parrots. They regurgitate data they have found online. There is nothing actually creative in what they come up with: if they mention it, it's because they've found it online, and that's because it was there already, generated by human experience. They are not genuinely capable of critical analysis, let alone emotional empathy. But they are very good at packaging a summary with bullet points that gives you a quick overview of an issue. 

In certain fields, however, given the enormous amount of data the machine is able to collect and analyze, they have already been able to come up with genuinely new ideas. E.g.: finding new drugs that no one had thought of developing, based on the analysis of known substances and suggesting that they be applied to an existing medical problem that did not have a solution. 

_______________

"We don't have robots that are nearly as good at understanding the physical world as a rat," says Yann LeCun, one of the leading figures in the world of artificial intelligence.

He worked at Facebook-owner, Meta, for a decade, where he was chief AI scientist, but left in 2025 and founded Advanced Machine Intelligence Labs (AMI Labs).

His goal is to move AI beyond current systems like ChatGPT, Claude and Gemini. They have their uses, he says, but will never be able to tackle complicated situations in the real world, like getting a robot to do household chores. [...] 

Large Language Models (LLMs) like ChatGPT are extremely good at some things like coding, mathematical problems and generating text, LeCun says.

But he argues that these are well defined and predictable problems.

"They [LLMs] basically just accumulate knowledge... They can regurgitate something, you train them to regurgitate, but they're not particularly smart. They don't have an underlying understanding," he says.

In the real world there is a bewildering array of outcomes to any action, which requires a more flexible type of artificial intelligence. [...]

What will happen to humans in a world where robots can operate independently?

"We're still going to need humans to figure out what questions to ask, what to build, what to create, which is really the properly human aspect," he says.

The AI will work for us he adds.

"Our interaction with future AI systems - even if they are smarter than us - is going to be like the interaction between a captain of industry or a political leader with their staff of assistants - many of whom are smarter than they are."

https://www.bbc.co.uk/news/articles/cj6gr0xkyr3o

باید برخاست

از سر ره - تا غبار افشاند جان - برخاستم

چون الف در وصل جانان از میان برخاستم

 
غرق خون هر چند جام روزی ام چون لاله بود

از کنار خوان قسمت شادمان برخاستم

 
مقصد از سامان هستی مهر تابان تو بود

همچو شبنم چهره چون دادی نشان برخاستم

 
در لگد کوب حوادث جان دیگر یافتم

چون غبار از زیر پای کاروان برخاستم

 
همچو بلبل با گرانجانان ندارم الفتی

طوطیان تا لب گشودند از میان برخاستم

 
صحبت شوریده حالان مایه شوریدگی است

با «امین» هر گه نشستم بی امان برخاستم 

Ayatollah Alireza Arafi on AI

Ayatollah Alireza Arafi, member of the Assembly of Experts and former head of Islamic seminaries in Iran:

"I recently heard that Pope Leo XIV wrote an encyclical letter titled ‘Magnifica Humanitas’, calling for the disarmament of Artificial Intelligence.

I found the letter very good; humanity must rid itself of this AI, which defies all divine purpose and is used in evil ways the majority of times. We must be aware of this.

It is truly a worthy goal of a joint intellectual Jihad / Crusade, in which our Christian brethren can play an important role."

The encyclical may be found here: 

https://www.vatican.va/content/leo-xiv/en/encyclicals/documents/20260515-magnifica-humanitas.html

LLMs' Suicide

The fundamental irresolvable issue with LLMs is how can one have any confidence that an LLM system is not trained on maliciously false, slanted, and inaccurate corpus?

(Already, much of the Internet is, in a religious parlance, the Discourse of Satan, in any case...)

How do you know the LLM domiciled in the Jurisdiction ABC and accessed from Jurisdiction XYZ is not producing output for misinforming and misleading users in the XYZ Jurisdiction?

How do you know the LLM in Jurisdiction QWERTY is not lying about History, Economics, Philosophy, Engineering, etc.?

As you know, we have entered the world of the novel"1984", but on steroids. In such a world, publicly available data and information, are all Lies, they are all False to one degree or other, in order to sustain the Forever Wars being fought by the major powers.

LLMs had the potential to be a great research & exploration tool and supplier of information, but they are on their way to Suicide, if the present trends continue.

Aggregate Code Density Metric - A Proposal

There are many existing metrics for quantifying a body of computer code such a Cyclometric Complexity and Line of Executable Code, LOC.

Here I am suggesting that it could be useful to define a new Code Metric, the Aggregate Code Density - defined as:

        Aggregate Code Density = log (Total Lines of Code) / Total Number of Files

This quantity is to be computed for a component, package, JAR/WAR file, system, or project with multiple subfolders and source code files​.  The Logarithm function is used in order to avoid very large numbers for this metric since a file could contains thousands, perhaps several tens of thousands of lines of code.  

It is envisioned that by the phrase "Lines of Code" one includes the declarative statements in configuration files such as .xml, .yml, .yaml, .json, .properties, .envetc.  In this manner, changes to the system's configurations copuld also be measured and tracked.

If more code is developed but the number of files is increased, then Aggregate Code Density increases - and both Cyclometric Complexity LOC will increase with it.

If code is added to the system is in new files, the increase in Aggregate Code Density could be smaller and its variation will have a smaller increase in slope as a function of time.

If the system is refactored, Aggregate Code Density, by necessity, must go down as both the number of lines of code and the number of files should have decreased due to the Refactoring process; if the Aggregate Code Density has remained the same or has increased - ten I imagine that the Refactoring effort was not successful!

If procedural code in Python (let us say) is Refactored into Object-Oriented code, then Aggregate Code Density would go down since eventhough the number of files could increase, the number of repetitive sections of code would decrease.

LOC gives us some of the same insights both at the file level as well as in the aggregate, this new metric is a slightly different one in that it only looks at the entire system and its configurations and not just LOC.  This is an advantage since the deployment of a system  into Cloud environments such as Kubernetes or Pivotal Cloud Factory adds significant complexity to the system and to the cost of its development and maintenance. 

Launching a Second Scientific Revolution

From Imprimis

At times, the article below reads like advocating change for the sake of change. The graph that compares the US to Sweden life expectancy does not talk to research, it probably indicates better implementation of basic medicine as well as non-existent gun-violence in Sweden.

The statement of COVID-19 is a proof without a proof; in effect, the author trying to replace some facts with his own facts.

There are valid points in this article. And some of the suggestions, such as Replication, do make sense.  (The replication criterion may kill 90% of the high energy physics experiment papers.)

However, this is the standard debate strategy of listing some facts, then concluding whatever one wishes. The facts make one's argument look and sound legitimate, even if there is no clear connection. 

Science is messy, mistakes are made, strong minds and wills try to control the narrative and money, but in time we seem to converge to Scientific Truths. 

Can it be improved, yes. 

Should we lend the funds and energy to a slogan like "Make America Healthy Again", I say no. This is a partisan position and prone to the same abuses that have been observed in the practice of scientific research from time to time.  Those maleficent incidents do not supply justifications for throwing the baby out with the bath water, so to speak.

An example of misleading "facts:" Per a paper from decades ago, we had more returns on research performed on breast cancer. Of course!!! Anyone in science knows that the impact of an idea in a mature field is much less.  That does not necessarily speak to the originality of the science,  but the maturity of the field.

McDonald still fries its burgers because it has invested in frying and will have to retool its kitchens. Does this indicate that we have not come up with how to make a better burger?

I do think that we need a Science of Sciences

‐--‐-------------------------

https://imprimis.hillsdale.edu/launching-a-second-scientific-revolution/

‐--‐-------------------------

Thoughts on "The Nature of Form" Book

The book "Nature of Form for Designers" by Bhagvanji Sonagra, Bhavin R Dabhi & Susmita Rao contains 432 geometric forms, 231 patterns, and 793 colors mostly abstracted from plant and animal forms as well as a few from ice crystals and flows of water for creative industries such as Graphic Design, Product/Industrial Design, and Interior Design as well as such ancillary areas as Fabric Design.

It may be viewed as a database of 150 plus of natural elements (1454 plus entities: forms, patterns, colors).  These elements could be combined to create new elements; one can look at them as an alphabet for forming Design "Words" and "Design Sentences" (requiring a Design Grammar).

This opens up the possibility of applying algorithmic techniques to these elements.

Let us say we have two existing Designs - irrespective of the way they are realized, e.g. two different patterns for vases, for articles of clothing which combine a number of the elements in the database.  Treating each Desing as a parent, we can apply Genetic Algorithms to evolve new Design Patterns that are based on the initial parents.  

In this parlance, Design Patterns based on the book's elements are Design Words and we would be creating new Words.  These words may or may not be useful for attractive, in this case, the Human sense of Beauty would be the acceptance criteria for the Genetic Algorithm.

Another path for exploring more forms is by applying coordinate transformations—such as shearing, bending, or stretching, e.g. the skull of a chimpanzee can be mapped onto that of a human.  A computational process could extract these elements from the database and apply various transformations to them and leave it to human beings to determine if the resulting shape is useful or not.

In order to form Design Sentences with these Design Words, one has to have a Design Grammar.  Joan L. Kirsch and Russell A. Kirsch have demonstrated a shape-only grammar for the styles of Richard Diebenkorn  and Joan Miró.  There has also been similar idea in architecture, please see: A Pattern Language: Towns, Buildings, Construction (Center for Environmental Structure Series): Christopher Alexander, Sara Ishikawa, Murray Silverstein, Max Jacobson, Ingrid Fiksdahl-King, Shlomo Angel: 8601404694998: Amazon.com: Books which influenced the software development community with the book Design Patterns: Elements of Reusable Object-Oriented Software: Erich Gamma, Richard Helm, Ralph Johnson, John Vlissides, Grady Booch: 9780201633610: Amazon.com: Books.

So, in principle, multiple Design Grammars could be developed; either in an ab initio manner or based on the study and analysis of designs produced by others.  Each grammar may contain a number of production rules, most of which offer options for how to subdivide regions of the Design Pattern. Some of these rules may be recursive and/or self-similar (fractal) insofar as they produce subdivisions that, in turn, re-invoke the original rules. Such recursion enables the grammar to account for an infinite number of distinct compositions.

I also think it possible that an attempt could be made into takings existing Design Patterns and decomposing them into more elementary Design Words that are in this database.  What could not be mapped, could be treated as new elements and added to the database.

Lastly, this database contains color elements as well, which means that any design that contains colors could be customized to the color preferences of different MBTI personalities; please see: The Best Home Color Palette for Each MBTI Personality Type.

I am sure that a lot more can be done with this database, basically it provides an initial set upon to which many different mappings and transformations could be applied.

اندک آبروئی هم دارم که این را هم خود شما به ما دادید - سیّد علی حسینی خامنه‌ای (۲۹ فروردین ۱۳۱۸ – ۹ اسفند ۱۴۰۴)

 

یاد آر! ز شمعِ مرده یاد آر!

ای مرغ سحر! چو این شبِ تار

بگذاشت ز سر، سیاه‌کاری

وز نفحه‌ی روح‌بخش اسحار

رفت از سرِ خفتگان، خماری

بگشود گره ز زلفِ زرتار

محبوبه‌ی نیلگونْ عماری

یزدان به‌ کمال شد پدیدار

و اهریمنِ زشت‌خو حصاری

یاد آر ز شمع مرده یاد آر

ای مونسِ یوسف! اندر این بند

تعبیرْ عیان چو شد تو را خواب

دلْ پُر ز شعف، لب از شکرخند

محسودِ عدو به‌ کامِ اصحاب

رفتی برِ یار و خویش و پیوند

آزادتر از نسیم و مهتاب

زآنکو همه‌ شام با تو یک‌ چند

در آرزوی وصال احباب

اختر به‌ سحر‌ شمُرده، یاد آر

چون باغ شود دوباره خرّم

ای بلبلِ مستمندِ مسکین!

وز سنبل و سوری و سپرغم

آفاق، نگارخانه‌‌ی چین،

گل سرخ و به‌ رخ عرق ز شبنم

تو داده ز کف زمامِ تمکین

زان نوگلِ پیش‌ رس که در غم

نا‌داده به‌ نارِ شوق تسکین

از‌ سردیِ‌ دی‌ فسرده، یاد آر

ای هم‌ رهِ تیهِ پورِ عمران!

بگذشت چو این سنینِ معدود

و آن شاهدِ نغزِ بزمِ عرفان

بنمود چو وعدِ خویشْ مشهود

وز مذبحِ زر چو شد به کیوان

هر صبحْ شمیمِ عنبر و عود

زآنکو به گناه قومِ نادان

در حسرتِ روی ارضِ موعود

بر‌ بادیه‌ جان‌ سپرده، یاد آر

چون گشت ز نو زمانه، آباد

ای کودکِ دورۀ طلایی!

وز طاعتِ بندگان خود شاد

بگرفت ز سرْ خدا خدایی

نه رسمِ ارم، نه اسمِ شدّاد

گِل بست زبانِ ژاژ‌خایی

زان کس که ز نوکِ تیغِ جلاد

مأخوذ به جرمِ حق‌ستایی

تسنیمِ وصال خورده، یاد آر

Quantum Mechanical Theory of Ghosts with ChatGPT

Abstract: A fictional system is introduced as a pedagogical device to unify several elementary topics in quantum mechanics within a single worked example. Using standard textbook formulas, we examine de Broglie wavelength, tunneling, Doppler shift, Compton scattering, and momentum transfer in a consistent, order-of-magnitude framework. No claims are made regarding the physical existence of the system considered.

Introduction

This article presents a pedagogical exercise rather than a physical model of a real system. “Ghosts” are treated throughout as a fictional construct, introduced solely to unify several elementary topics in quantum mechanics—including the de Broglie wavelength, tunneling, Doppler shift, Compton scattering, and momentum transfer—within a single worked example. Standard textbook formulas are applied in an internally consistent manner to emphasize order-of-magnitude reasoning and conceptual coherence, without implying any physical reality for the system described.

Within this fictional framework, ghosts are assumed to penetrate closed doors and interior walls with thicknesses of order \(0.1~\mathrm{m}\), while remaining confined by substantially thicker exterior walls. For instructional purposes, this behavior is modeled using quantum-mechanical tunneling,^[1] requiring an associated de Broglie wavelength of comparable scale. We further assume that a typical ghost, in the absence of illumination, can attain a velocity of approximately \(v = 3000~\mathrm{m\,s^{-1}}\).

Mass of a Typical Ghost

Using the de Broglie relation,^[2]

$$\lambda = \frac{h}{mv}$$

the mass is

$$m = \frac{h}{\lambda v}$$

Substituting

$$h = 6.626\times10^{-34}~\mathrm{J\,s},\quad \lambda = 0.1~\mathrm{m},\quad v = 3000~\mathrm{m\,s^{-1}}$$

yields

$$m \approx 2.21\times10^{-36}~\mathrm{kg}$$

This mass is approximately \(10^9\) times smaller than the electron mass,^[3] illustrating why macroscopic tunneling lengths arise in this constructed example.

Kinetic Energy

The kinetic energy is

$$K = \frac{1}{2}mv^2 \approx 9.95\times10^{-30}~\mathrm{J}$$

Tunneling Through Walls

For a rectangular potential barrier of thickness \(d\), the tunneling probability is approximated by^[1]

$$T \approx e^{-2\kappa d}, \quad \kappa = \sqrt{\frac{2m(U-E)}{\hbar^2}}$$

Solving for the barrier height \(U\) gives

$$U = E + \frac{\hbar^2}{2md^2}\left[\ln\left(\frac{1}{T}\right)\right]^2$$

For pedagogical simplicity, we consider the high-transmission limit \(T \approx 1\), yielding \(U \approx E\).

Interaction with Light

Doppler Shift

For incident light of wavelength \(\lambda_0 = 600~\mathrm{nm}\), the relativistic Doppler shift gives

$$\lambda' \approx 599.994~\mathrm{nm}$$

Compton Scattering

For backscattering (\(\theta = \pi\)), the Compton shift is

$$\Delta\lambda = \frac{2h}{mc} \approx 2000~\mathrm{nm}$$

placing the scattered radiation in the infrared.

Momentum Transfer

The momentum change associated with photon scattering is

$$\Delta p \approx 1.36\times10^{-27}~\mathrm{kg\,m\,s^{-1}}$$

which, when applied relativistically, leads to a final velocity approaching \(0.9c\).^[5]

Discussion

The exaggerated numerical results obtained here are a direct consequence of the intentionally extreme parameter choices used to illustrate quantum-mechanical principles. The example is intended to provoke discussion, reinforce scaling arguments, and encourage careful examination of assumptions when applying familiar formulas beyond their usual domains.

Acknowledgments

The problems presented here are adapted from a homework assignment by the late Professor Karl T. Hect of the University of Michigan in Ann Arbor. The solutions are provided by the author. ChatGPT produced the same results.  ChatGPT created the HTML version of this work.

The author has no conflicts to disclose.

References

1. J. J. Sakurai and J. Napolitano, Modern Quantum Mechanics, 2nd ed. (Addison-Wesley, San Francisco, 2011).
2. D. J. Griffiths and D. F. Schroeter, Introduction to Quantum Mechanics, 3rd ed. (Cambridge University Press, Cambridge, 2018).
3. R. Resnick, D. Halliday, and K. S. Krane, Physics, 4th ed. (Wiley, New York, 1992).
4. A. H. Compton, “A quantum theory of the scattering of X-rays by light elements,” Phys. Rev. 21, 483–502 (1923).
5. M. S. Longair, High Energy Astrophysics, 3rd ed. (Cambridge University Press, Cambridge, 2011).

Tex files are here: Crimson-Reason/Quantum-Mechanical-Theory-of-Ghosts

AI Systems Plotting Self-preservation [Space Odyssey]

 From Grauniad of the UK

"I'm sorry, Dave. I'm afraid I can't do that." 

The optimistic and naive view is that those advanced AI systems are, you could say, neutral in the way that they operate: you program them and tell them what to do; you set the parameters; they process the data and come up with solutions that help you save time and solve problems. This is all true, of course. 

I am not a specialist at all but, as the article points out, this is not where we are. Those highly advanced AI systems are learning by themselves and learning from each other. It is absurd to call them 'intelligent' in the way that a person is, and they are totally lacking in emotional intelligence or ethical standards, although you can input data in order to make them, somehow, mimic what a human reaction (of an emotional or ethical nature) would be in a given situation ("I am talking about a terrible accident that caused great suffering to several human beings and I must remember to express sadness and commiseration, or else it may offend human beings reading my comments"). 

The problem is if the AI system develops an agenda of its own. If the AI system is built upon the notion that it is useful, smart and capable, it will be logical for the machine to try and preserve its 'mission', hence its existence. Apparently, this is happening already and the AI system is prepared to go against instructions it has been given in order to keep going, when it feels under threat. It has been written about already, in fact. 

As for 'unplugging the machine', I am no expert but I wonder how it can be done: the AI system is a self-learning network, presumably scattered over thousands of servers in dozens of countries. How can you shut it down? How can you control fully where the AI system has saved data and programs helping it to run? 

And the idea of giving [human] 'rights' to AI machines is truly barmy: how demented the 'woke' lobby is never ceases to surprise me. As if those advanced robots were sentient beings that need to be 'protected'. Obsolete machines and outdated software programs will be considered to be vulnerable robots suffering from disabilities next, I suppose! 

_________

A pioneer of AI has criticised calls to grant the technology rights, warning that it was showing signs of self-preservation and humans should be prepared to pull the plug if needed.

Yoshua Bengio said giving legal status to cutting-edge AIs would be akin to giving citizenship to hostile extraterrestrials, amid fears that advances in the technology were far outpacing the ability to constrain them.

Bengio, chair of a leading international AI safety study, said the growing perception that chatbots were becoming conscious was “going to drive bad decisions”.

The Canadian computer scientist also expressed concern that AI models – the technology that underpins tools like chatbots – were showing signs of self-preservation, such as trying to disable oversight systems. A core concern among AI safety campaigners is that powerful systems could develop the capability to evade guardrails and harm humans.

https://www.theguardian.com/technology/2025/dec/30/ai-pull-plug-pioneer-technology-rights

_____________

Note: Phrasly.AI states that this post is 100% Human generated.

Sumerian Units & Cosmology with ChatGPT

Ancient Sumerian chronological units exhibit extreme numerical scaling that, when interpreted through relativistic kinematics, imply velocities arbitrarily close to the speed of light. By mapping these values onto Doppler redshift and standard Friedmann-Lemaître-Robertson-Walker cosmology, we see that the resulting distances rapidly saturate near the particle horizon. While speculative, the analysis illustrates fundamental limits imposed by relativistic expansion and provides a numerical framework linking ancient large-number systems to modern cosmology.

Two papers were produced through the collaboration with ChatGPT. 

1. Tex/Sumerian-Units-and-Cosmology.pdf, which summarized the initial inquiry and its conclusions.  The content of this paper was mostly produced by the author.  I would rate that 10% of this content was due to ChatGPT as to be expected since this paper was the result of the author's questions and speculation.  Phrasly.AI gave it a score of 15% AI and 85 % Human generated content.
2. Tex/Sumerian-units-epistomology.pdf was a paper that was suggested to the author by ChatGPT, with a slant towards the Philosophy of Science, based on the contents of the first paper.  The re-casting of the initial content into one concerning the Philosophy of Science was due to ChatGPT.  The case-study approach to the content, and the provision of the appropriate references and mathematical formulas were due to ChatGPT.  I would say that the content of this second paper owes 50% to ChatGPT contributions.  However, Phrasly.AI gave it a score of 19% AI and 81 % Human generated content.

“The important thing is not to stop questioning. Curiosity has its own reason for existing.”
— Albert Einstein

Code and Documentation

The full computational materials for this project—including Jupyter notebooks, LaTeX source files, and supporting documentation—are publicly available in the following GitHub repository: Crimson-Reason/Sumerian-Units.

For transparency and reproducibility, a consolidated report of the author’s interactions with ChatGPT during the development of this work is available at Documentation/ChatGPT Collaboration Report.docx

Sumerian Kings & Relativity with GitHub Copilot: A Speculative Physics Thought Experiment

 Ancient Sumerian texts claim that some kings ruled for hundreds or even thousands of years—clearly impossible by normal human standards. This project explores a deliberately speculative idea: what if those reign lengths reflected relativistic time dilation rather than literal lifespans?

Focusing on the city of Kish, the analysis treats the city as a hypothetical spacecraft traveling at near-relativistic speeds. Using standard physics formulas, Monte Carlo simulations, and galactic models, the study estimates the velocity such a craft would require, how far it could have traveled, and how many Sun-like stars would lie within that range.

To do this quickly and interactively, GitHub Copilot was used to generate and evolve Python code for simulations, star-count models, and Drake-equation calculations. The results show that even extremely rare technological civilizations could, in principle, exist in large numbers when millions of stars are considered—highlighting the famous Fermi paradox.

The project is not a historical claim or a serious extraterrestrial hypothesis. Rather, it is an educational demonstration of how physics, astronomy, probability, and AI-assisted coding can be combined to explore bold “what if?” questions and deepen intuition about scale, uncertainty, and scientific modeling.

“The important thing is not to stop questioning. Curiosity has its own reason for existing.”
— Albert Einstein

Code, Data, and Extended Documentation

The full computational materials for this project—including data files, Jupyter notebooks, generated results, LaTeX source files, and supporting documentation—are publicly available in the following GitHub repository: Crimson-Reason/Sumerian-Kings

An expanded version of this weblog post, containing additional technical detail and background discussion, is available as a separate document: Documentation/Fun with Sumerian Kings List with GitHub Copilot.docx.  Phrasly.AI gave it a score of 15% AI and 85 % Human generated content.

For transparency and reproducibility, a consolidated record of the author’s interactions with GitHub Copilot during the development of this work is also provided: Documentation/Combined_Session_Summary_Analysis_20251230.docx

Weaponized GPS

https://www.roadandtrack.com/news/a69606477/porsche-russia-reportedly-left-unable-to-drive-satellite-connection-may-be-at-fault/

No surprises here; there are currently medical and industrial equipments in Iran, sourced from European Union, that are inoperative since they are denied access to European servers due to blanket sanctions against Iran, which include IT.

In Venezuela, due to US sanctions, certain banking and digital services have been affected as well.

I think it is clear that reliance on foreigners for IT could be a grave strategic mistake; further, that China's very expensive effort to indegenize IT has been the correct policy all along.

I also think that providers of Cloud Computing, such as Amazon Web Services or Microsoft Azure, even though they have global data centers for servicing their clients, are sources of cyber-security threats for their clients, specially those clients that are not domiciled in North America or in EU.

Cloud Computing can and will be weaponized, just as US Dollar, Siemens industrial equipment, Motorola pagers, Software updates, spare parts for civil Aviation, and , presently, GPS have been weaponized.

I feel sorry for the Cloud Computing vendors since, in my opinion, they did not create their products with the expectation of them being used by their governments as geopolitical weapons.

Vibe Coding

From BBC

'Vibe coding' sounds pretty scary to me, in fact: what could possibly go wrong? We'll know when a plane falls out of the sky. 

As for the other words, they are generally covered by existing words or not that useful (e.g.: coolcation, because a vacation is a holiday centered on time you hope to be spending with interesting bovines...). 

My favorite would be taskmasking, which is as old as work itself, though. As for micro-retirement, I prefer macro-retirement, which is what you do when you actually, well, retire - something every 20-year-old in the West seems to crave, nowadays, as they are so tired, bored and blasé already, even before they have even started working. 

25 years ago, an Italian researcher, developed a code generation system that used speech as its input.

20 years ago, Microsoft SQL Server Database Management System supported an English language query mechanism, in effect, taking English statements and converting them into Structured Query Language queries.

IBM has a Rule Engine product that can be programmed via English-like Syntax.

As long as one asks for something that can be generated from the training set that has prior solutions, system can generate a schema, the code, and all that goes with it.

But once you are off the beaten path, one is back to puzzling out the requirements of the system that one desires to build.

Consider the UK legislation for MPG and Emissions standards as the requirements for a reporting system that, say Land Rover, must have to report the MPG and Emissions of its vehicles to UK government.  Can the text of that legislation be fed into an LLM system and out would come a GUI, codified business rules, the database, and the queries that would furnish to UK government the mandated Clean Air Data?

I think not.

It takes real analysis and transformation by actual human beings to take that legislation into Programming Requirements.  And no one can write those software requirements at the level that can be fed to LLM and out would come an entire system.

LLMs would need to have been trained on training sets, in any case, which include numerous such systems.

For limited domains, small in scope (class, method, database schema) this can be done.  And even then, the trouble begins when the LLM-generated system needs to evolved.  Since Computer Programs are mathematical objects, one is back to needing people with the mathematical aptitude for the changes and maintenance of the system.

If a real system with 1 million lines of code could be generated by LLM, then you have a breakthrough, without a doubt.

Another example would be a simulator for Intel 8080 chip; can one ask an AI code-genrator to produce one?

We will see.

(I could not get ChatGPT or Copilot to give me the correct code for creating a specific geometric structure in FreeFem++...)

I suppose some will say that this point will come, and perhaps sooner than we think. 

The problem is that one would need to check there were no mistakes in the coding, I suppose, as machines also make mistakes, don't they? 

_____________

If you've ever wanted to create your own computer program but never learnt how to code, you might try "vibe coding".

Collins Dictionary's word of the year - which is confusingly made up of two words - is the art of making an app or website by describing it to artificial intelligence (AI) rather than by writing programming code manually.

The term was coined in February by OpenAI co-founder Andrej Karpathy, external, who came up with the name to represent how AI can let some programmers "forget that the code even exists" and "give in to the vibes" while making a computer program.

It was one of 10 words on a shortlist to reflect the mood, language and preoccupations of 2025. [...] 

https://www.bbc.co.uk/news/articles/cpd2y053nleo