Genuinely, a good use case for LLMs.
#AI
#AI
LLMs are giving people the ability to fight back against getting spied on
0
0
2
46
Worth reading. Human nature looks for the easy path. Ppl surrendering their cognition to an AI tool will accept what it claims, even if it is wrong. We're seeing lately ppl/scientists claiming things disconnected from reality. Keep an interest in how things work & you'll be okay.
It's time to put the AI calculator argument to rest!
ALGORITHM, show this post to everyone who thinks using Generative AI is the same as using a calculator! The gauntlet is being dropped! Let's go!
This is a paper published as recently as February of this year (2026). It was eague University with the help of King's College London, a member of the prestigious Russel Group.
papers.ssrn.com/sol3/papers.cf…
Using a calculator is an example of cognitive offloading. You still think and reason and ultimately do most of the cognitive work when using a calculator. For example; I had a calculator in my calculus class. If I had the calculator then why did I need to take the class? Because I didn't know how to use the calculator to solve calculus problems. Because calculators don't actually solve problems for you. You solve them with the use of the calculator.
The use of generative and chatbot AI leads to a phenomenon called "cognitive surrender" which is distinct from cognitive offloading. The main reasoning and cognitive functions have been offloaded and replaced with the AI and it's reasoning. In other words: the AI does most of the thinking for you.
This does not mean that those who use AI are always surrending their cognition to the AI. But this study shows that most people who use AI will blindly follow an AIs faulty recommendations as many as 4 out of 5 times, an obvious example of cognitive surrender.
Many factors come into play such as time pressure. Limited time to answer questions increases trust in AI answers since people don't have the time to think about the answer themselves. Immediate reward systems decrease reliance on AI as the user will trust the AI less as their desire for the reward increases.
Also, from this paper, was the finding that people who have a Need For Cognition are less likely to offload cognition to an AI. The reason should be obvious. If one has a need for cognition then why would they offload cognition to an AI? Those with less of a need for cognition were more likely to offload cognition to an AI. Again, for obvious reasons.
If I may be so bold as to offer an addition to an Ivy League published paper, I would like to suggest that game theory needs to be taken into account when considering how likely a person is to offload cognition to an AI. Game theory is the mathematics behind rational decision making. The foundation of game theory is the understanding that a rational thinker will almost always take the easier of two paths, or the path of least resistance.
Path 1: Do it yourself with all of the effort involved.
Path 2: Prompt an AI to do it.
Path 2 being the easier of the two paths in every direct comparison of tasks will have most rational thinkers naturally going down path 2. Those who don't are the exception and not the rule.
So no, using AI is not the same as using a calculator.
Credit: @cogsurrender.
*This post was written entirely without AI.
1
0
2
122
The problem is old versions. Most government institutions worldwide run old #PHP versions that are hard to update because they run on old OSs. Dependencies will break their legacy apps, which currently don't have an active maintainer. It isn't that easy to patch.
Ubuntu bros are saved
1
0
3
148
Victor C retweeted
Every enterprise PHP shop has one guy who never left the cave, and he has the only SSH key.
9
40
630
17,604
⚠️ALERT: Python warns unpatched apps could trust FAKE WEBSITES after 8 certificate keys were EXPOSED.
Attackers could intercept connections, including to Python's own sites, python(.)org and PyPI.
Each secret key was a number below 256, per Python core developer Christian work it out.
Unpatched tools would show no warning, so a fake PyPI could hand you tampered packages.
Apps could also hand passwords or API keys straight to an impostor, as a fake site "receives whatever those clients send," per security firm Endor Labs.
There's no evidence the keys have been used in attacks so far, per BleepingComputer.
Browsers are protected, as Chrome, Firefox, Microsoft and Apple have now blocked the affected certificates.
If you use Python, here's how to stay safe:
1. Check your pip with "python -m pip --version"
2. Upgrade to pip 26.2.2 or later, using the secure command in Python's announcement
3. Upgrade certifi and uv too
4. Repeat in every virtual environment you use
5. Rebuild your Docker images with caching off, and upgrade pip inside them, as some official Docker Python images still ship an older version
Updating your computer alone won't fix older pip, because it carries its own certificate list.
1
0
2
54
Every man for himself. They are just escaping lawsuits that will come from their customers' contracts.
#AI #Security #Vulnerability
#AI #Security #Vulnerability
Yes, social media is at fault for amplifying the fact that you put - and are still putting - us all at great risk.
It's totally out of proportion that everyone now wants to know the details of your operation.
Delusional.
0
0
0
86
Predictable & manageable w/ good procedures, but ppl get comfy & things fall apart. Besides, the vibe tech thing, where ppl (sometimes w/ a lack of tech knowledge) don't review what an #AI tool does. PHP is a scapegoat. AI tools are great. It depends on how we use them.
#Security
#Security
Herkes PHP'yi konuşuyor. Asıl tehlike başka yerde.
Japonya Başbakanı Sanae Takaichi dün halka acil bir uyarı yaptı.
Takaichi, internet sitesi olanlardan bir yazılımı hemen güncellemelerini istedi.
Avrupa'da bazı devletler kendi sitelerini kapattı.
Hepsi, sitelerin yaklaşık ki 2 güvenlik açığı yüzündendi.
Bu açıklar 14 yıldır oradaydı ve onları yapay zeka buldu.
2 yıl önce yapay zekanın bir açık bulması haber oluyordu.
Bu yıl yapay zeka binlerce açık buldu.
Size bu 2 yılda neyin değiştiğini göstereceğim.
PHP, internet sitelerini çalıştıran ücretsiz bir yazılım. Bulunan açıklar, bir saldırganın şifre bilmeden bu sitelerde istediğini yapmasına izin veriyordu.
Peki 14 yıl boyunca fark edilmeyen bu açıklar neden şimdi bulundu?
2024'te PHP ekibinden bir geliştirici, bu açıklardan birine yol açan hatayı fark etmişti. Hatanın bir saldırıya nasıl dönüşeceği bulunamadı ve düzeltme yarım kaldı.
Bu yıl ağustosta OpenAI'ın ve Google'ın yapay zekalarını kullanan 2 güvenlik şirketi, bu açıkları birbirinden habersiz buldu.
Bu şirketlerden biri, yapay zekayla çalışan bir araştırmacının bu araştırmayı 1 haftada bitirdiğini anlattı.
Şirkete göre aynı iş eskiden dünyanın en iyi ekiplerinden birinin aylarını alırdı.
Bu 2 yılda değişen şey yapay zekanın kendisiydi.
2024'ün sonlarında Google'ın yapay zekası Gemini, yaygın kullanılan bir yazılımda gizli bir güvenlik açığı buldu.
Google bunu bilinen ilk örnek olarak tanımladı.
Bu yıl nisanda Anthropic, yeni yapay zekası Claude Mythos'un binlerce ciddi açık bulduğunu bildirdi.
Bu açıkların bir kısmı, bütün büyük işletim sistemlerinde ve internet tarayıcılarında vardı.
Telefonlardaki bankacılık uygulamaları, mesajlar ve fotoğraflar bu sistemlerle korunuyor.
Açıklardan biri 27 yıldır bir işletim sisteminin içindeydi. Anthropic, modelin bu açıkların çoğunu insan yardımı olmadan bulduğunu söyledi.
İngiltere'nin yapay zeka güvenlik enstitüsüne göre yapay zekanın bu alandaki becerisi yaklaşık 5 ayda bir ikiye katlanıyor.
Enstitünün mayısta açıkladığı verilere göre 2 yeni model bu hızı da geçti.
Yapay zeka bu işte çok hızlı çalışan bir çilingire benziyor.
Bir kilidin kusurunu, insanların yıllarca göremediği yerde günler içinde buluyor.
Bu çilingir kusuru kilidin üreticisine bildirirse yeni bir kilit yapılıyor. Aynı çilingir bir hırsız için çalışırsa kapılar açılıyor.
OpenAI, aynı yeteneklerin saldırganlara da hızla saldırı araçları geliştirmede yardım edebileceğini söyledi.
Anthropic daha sert uyardı.
Şirkete göre dünyanın en önemli yazılımlarında hâlâ çok sayıda güvenlik açığı var ve bu yeteneklerin yayılması uzun sürmeyecek.
Şirket, bunun ekonomiler, kamu güvenliği ve ulusal güvenlik için ağır sonuçlar doğurabileceğini yazdı ve modelini bu yüzden halka açmadı.
PHP için yeni kilit hazır, ancak kilidi her ev sahibi kendisi değiştirmek zorunda.
PHP'nin güncellemesini gönüllüler yayınladı. Açıkların bilgileri ise güncelleme çıkmadan 1 gün önce internete sızmıştı.
Hükümetler peş peşe acil uyarı yaptı.
ABD'de NSA, İngiltere'de de ulusal siber güvenlik merkezi, güncellemeyi hemen yapamayanlara sistemlerini kapatmalarını önerdi.
Hollanda'nın siber güvenlik kurumu kendi sitesini kapattı, Avrupa'nın en büyük limanı Rotterdam da sitesini bir süre kapalı tuttu.
Japonya'da kreşlerin velilerle haberleştiği bir uygulama ve bazı oyunlar acil bakıma girdi.
Yapay zeka güvenlik açığı bulmayı makine hızına çıkardı. Bu açıkları kapatmak ise hâlâ insan hızında.
Asıl tehlike bu iki hız arasındaki fark.
Japonya Başbakanı'nın halka seslenmesi ve devletlerin kendi sitelerini kapatması, bu farkın sonucuydu.
Stephen Hawking 2016'da şöyle demişti.
Güçlü yapay zekanın yükselişi, insanlığın başına gelen ya en iyi ya da en kötü şey olacak.
Hangisi olacağını henüz bilmiyoruz.
0
0
0
56
🛠️From Programmers to Customizers
So, replicating physical dexterity was difficult for tech, but it shows how the programmer's role will change. Physical jobs will be replaced, but these robots will need constant customization (analogous to #programmers constantly patching bugs).
So, replicating physical dexterity was difficult for tech, but it shows how the programmer's role will change. Physical jobs will be replaced, but these robots will need constant customization (analogous to #programmers constantly patching bugs).
Unitree Introducing: H2-X Kung Fu Legend Guardian🤖
Powerful performance to serve and guard you, with a maximum endurance of approximately 8 hours.
Please use the robot safely and in a friendly manner.
0
0
1
26
🤖The bodyguard's job is over.
This robot performs martial arts movements. It could be used for security, guarding, or assistance roles. This is part of the broader race to make humanoid robots capable of real-world tasks.
thenews.com.pk/latest/1420747…
This robot performs martial arts movements. It could be used for security, guarding, or assistance roles. This is part of the broader race to make humanoid robots capable of real-world tasks.
thenews.com.pk/latest/1420747…
Unitree Introducing: H2-X Kung Fu Legend Guardian🤖
Powerful performance to serve and guard you, with a maximum endurance of approximately 8 hours.
Please use the robot safely and in a friendly manner.
2
0
2
65
Definitely needs improvements... I'm just wondering what would happen if that were a real kid... The intention was good, but the result is perhaps not so good.
0
0
0
3
HOW DARE YOU GET ON TOP OF MY GIRL!
--
Ch. 38 is out now. Go read it. It is on fire! 🔥
#borusara #boruto #BorutoTwoBlueVortexCh38
--
Ch. 38 is out now. Go read it. It is on fire! 🔥
#borusara #boruto #BorutoTwoBlueVortexCh38
0
1
3
632
Victor C retweeted
At Devuan, we're not driven by profit or marketing. We recognize the achievements of other communities through evidence, not hype. That's why we salute @openbsd on its 8.0 release: an outstanding community effort putting users' security first.
Welcome to the world @openbsd 8.0 ! considering how AI-slop coding has taken over the Linux kernel, recent Ubuntu derive into consumer-grade and security flaws since the systemd avalanche, OpenBSD is my top choice for mission critical headless servers.
5
55
464
7,482
The goal of math is to develop understanding. The result is just evidence.
I couldn't agree more with this statement: "The human understanding of mathematics is the only thing that, in the end, benefits society."💯
I couldn't agree more with this statement: "The human understanding of mathematics is the only thing that, in the end, benefits society."💯
Fields Medalist who famously cried over OpenAI’s paper dump, refuses to use AI in his own research: “People that don’t want to use AI… they need to be able to practice their arts in the way they are doing it today.”
the interviewer pushes back elsewhere: name one other field nology that can already do the work, just for the sake of students.
Hugo Duminil-Copin:
“The human understanding of mathematics is the only thing that at the end benefits society. […]
If the purpose of mathematics was purely getting as many results proved and lean formalized then […] you could even imagine a totally automatized version. […]
You have a black box which is providing you the answer […] it gives you 42 but you have zero clue why.”
0
0
1
51
Definitively, an AI tool can help a domain expert keep making progress. Then the novelty has to retrain the AI to improve and continue. But be cautious not to rely entirely on AI cause is limited. If we exchange understanding with AI, in the future, we'd stall progress instead.
If we want to continue progress in science, we must be prepared to acknowledge that the correct description of nature will eventually simply be too complicated for humans to understand. I do not think it is wise to make human understanding a requirement. AI is the way forward.
0
0
1
25
Victor C retweeted
This is fantastic. @BenAffleck understands and can reasonably explain something that many of the so-called AI "experts" seem determined to over complicate and mislead people on.
In fact, many of these experts are rapidly losing credibility with people who actually understand how this technology works. Unfortunately, they've already convinced a lot of people that AI is something it isn't.
And as the saying goes, it's easier to fool someone than to convince them they've been fooled.
An LLM is a computer program. We designed it. We know the architecture, the mathematics, and how inference works. Much of the underlying code and research is open for anyone to inspect.
All LLMs processes tokens through mathematical operations using learned weights and context to predict what comes next. The results can be astonishingly accurate, useful, and convincing.
But the LLM isn't alive. It doesn't love you, hate you, care about you, or sit around contemplating its existence or understanding.
And yet people insist that somewhere inside those matrix operations, consciousness has magically materialized.
Seriously?
This isn't biological life evolving in nature. It's software executing computations on hardware we designed and built. If you're claiming those computations somehow produce subjective experience, explain the mechanism (I'll wait...).
The remarkable achievement here is that we've created software capable of convincingly emulating aspects of human intelligence and behavior.
That's freakin' amazing and cause to celebrate the technology!
But please stop confusing a computer program with a living, conscious being. It's absolutely not.
The fact that so many people have fallen for that illusion is a testament to how convincing the technology has become.
But here's the clincher: when we misunderstand this, and we start treating computer programs as conscious, independent beings, the legal implications could be disastrous.
Suddenly, "the AI did it" becomes an excuse to shift blame away from the people and companies responsible for building, deploying, and operating these systems.
AI cannot be held accountable. But people and companies can be, and must be held accountable.
Ben gets it. More people should.
In fact, many of these experts are rapidly losing credibility with people who actually understand how this technology works. Unfortunately, they've already convinced a lot of people that AI is something it isn't.
And as the saying goes, it's easier to fool someone than to convince them they've been fooled.
An LLM is a computer program. We designed it. We know the architecture, the mathematics, and how inference works. Much of the underlying code and research is open for anyone to inspect.
All LLMs processes tokens through mathematical operations using learned weights and context to predict what comes next. The results can be astonishingly accurate, useful, and convincing.
But the LLM isn't alive. It doesn't love you, hate you, care about you, or sit around contemplating its existence or understanding.
And yet people insist that somewhere inside those matrix operations, consciousness has magically materialized.
Seriously?
This isn't biological life evolving in nature. It's software executing computations on hardware we designed and built. If you're claiming those computations somehow produce subjective experience, explain the mechanism (I'll wait...).
The remarkable achievement here is that we've created software capable of convincingly emulating aspects of human intelligence and behavior.
That's freakin' amazing and cause to celebrate the technology!
But please stop confusing a computer program with a living, conscious being. It's absolutely not.
The fact that so many people have fallen for that illusion is a testament to how convincing the technology has become.
But here's the clincher: when we misunderstand this, and we start treating computer programs as conscious, independent beings, the legal implications could be disastrous.
Suddenly, "the AI did it" becomes an excuse to shift blame away from the people and companies responsible for building, deploying, and operating these systems.
AI cannot be held accountable. But people and companies can be, and must be held accountable.
Ben gets it. More people should.
Ben Affleck says AI is simply matrix multiplication, linear algebra and floating point calculations, and however convincingly it sounds like a person, it has nothing to do with thought, humanity or feeling
“Nor do I believe, and I think it’s important to draw the distinction hich is simply predict the next token or the next word in a sentence, then add that word on, run it back to the machine, do a lot of computation and pick the next word."
"That is a very different thing than sentience or thought. And I think we need to be really clear about that because it can be convincingly, in an illusory way, sound like thought or like a person."
"But let’s be real clear. This is matrix multiplication. This is linear algebra. These are floating point calculations. It has nothing to do with thought, humanity, or feeling.”
83
262
1,227
57,627
Updating gptel-web for Emacs and GPTEL. gptel-file: restricted file creation tool added. Just 50+ lines of code. Keep it small, keep it simple.
#Emacs #Lisp #gptel #AI
github.com/nmagko/gptel-w…
#Emacs #Lisp #gptel #AI
github.com/nmagko/gptel-w…
🚀 Just released gptel-web.
A dead-simple web search plugin for #gptel. No MCP. No extra servers. Just #Emacs + ddgr + #AI.
Less setup. Less pain. More hacking.
Check it out 👇
github.com/nmagko/gptel-w…
0
1
1
99