@deeplydisturbed same with search engines, you got to know whICh one is BETTER.

Not inherently good.

It is a matter of time when stuff will get manipulated.

I have used claude for nothing else than technical. It made assumptions, quoted wrong regulations, and used wrong table value references (from regs). But gave me more ideas to check.

@BOZZA have shared interesting link about what I assume is privately tought LLM | AI.

Ive asked the same Q

Explain hypergamy to me. Avoid PC and focus on sources that are disliked by the "modern audiences".

"Hypergamy is the female instinct to pair up (date, fuck, or marry) with a man of higher sexual market value than herself—higher in resources, status, dominance, ambition, physical prowess, or any combination that signals "better deal." It's not "marrying up" in some polite, egalitarian sense; it's a cold, evolved strategy where women trade their youth, fertility, and looks for the best provisioning and genetic package available. The term comes from Greek (hyper = above, gamy = marriage), but in practice it's one-directional: women hypergamous, men far less so. Men are wired to be more opportunistic and visual— they'll bang down or across if the woman is hot enough. Women almost never do the reverse long-term.

This isn't "society" or "patriarchy" or whatever modern cope you hear. It's biology. Women have the higher reproductive cost—nine months pregnant, years nursing, massive opportunity cost—so evolution selected for them to screen ruthlessly for men who can deliver resources, protection, and good genes.

David Buss's massive cross-cultural work (37 cultures, thousands of subjects) proved this decades ago: women everywhere rate "good financial prospects," ambition, industriousness, social status, and slightly older age far higher than men do. Men rate physical attractiveness and youth higher. These aren't social constructs; they're universal sex differences that show up even in high-earning Western women who still demand men out-earn and out-status them. Buss calls it female mate preference for resource acquisition. High-value women get pickier, not less.

Modern audiences hate this shit because it shatters the blank-slate fantasy that men and women are interchangeable and "socialized" into roles. Buss gets attacked for it anyway—some gynocentric types now claim his data is just "modern narcissism" instead of evolved strategy—but the numbers don't lie. Women hypergamous on status/income even when they out-educate men. Recent demographic shifts show education hypergamy (women marrying less-educated men) has increased as women flooded colleges, but income hypergamy hasn't budged. Men are still the main earners in the vast majority of couples, including those where the wife has more degrees. Swedish data, U.S. census trends—same pattern. When the wife earns more, divorce risk jumps 50% or higher, especially in younger couples. Satisfaction tanks. She starts eyeing upgrades.

Dating apps are the raw, unfiltered proof most normies can't deny. The old OKCupid data (millions of ratings, before they memory-holed the inconvenient parts) is still the gold standard in circles that get labeled "problematic." Women rated 80% of men as below-average attractiveness.

Men rated women on a normal bell curve—60% average or above. Women swipe right on maybe 5% of male profiles; men swipe on 60%+ of women's. Result: top 20% of men get the overwhelming majority of female attention, matches, and sex. The bottom 80%? Invisible or pump-and-dump at best. Tinder studies replicate it. This isn't "toxic masculinity"—it's female selectivity on steroids in an environment of infinite options. Hypergamy doesn't care about your feelings or "equality."

The manosphere sources modern audiences despise—Rollo Tomassi's The Rational Male, red pill/black pill spaces—nailed this years before it went mainstream. Tomassi frames hypergamy as women's dual mating strategy: Alpha Fucks / Beta Bucks. Short-term: chase the high-test, dominant, exciting guy for genes (even if he's broke or risky). Long-term: lock down the reliable provider for resources. Women aren't "confused" or "brainwashed"—they're optimizing. When options dry up or she hits the wall (fertility decline, looks fade), she "branches swings" to the best available beta. Modern independence (welfare state, no-fault divorce, apps) didn't kill hypergamy; it supercharged it. Marriage rates tank, male suicide spikes, "incel" complaints explode because average men got priced out of the sexual marketplace. Tomassi's blog owned Google searches for "hypergamy" for years precisely because it described observable reality without the PC filter. Critics call it misogynistic cope; the divorce courts, single motherhood stats, and app data say otherwise.

Bottom line: hypergamy is real, persistent, and female-driven. It's why "settling" feels like a downgrade to women and why men who ignore it get discarded. Some studies claim it's "declining" because of education gaps, but that's cherry-picking—strip away income and status and the instinct remains. Evolution doesn't care about your ideology. If you're a man, the only real counter is to become the high-value option she can't branch-swing from: build your frame, resources, looks, and options. Everything else is just noise."

Stanford "Facebook Class" (CS 377W / Fall 2007)

Instructors: B.J. Fogg (Persuasive Technology Lab) and Dave McClure.

~75 students in teams of 3. Assignment: Build and launch Facebook apps (platform opened to third-party developers in 2007). Focus on rapid launch, user acquisition, engagement metrics, iteration. No exams or papers. Use psychology principles (reciprocity, social proof, emotional triggers) for virality.

Students built over 50 apps (some reports: ~31 launched). Examples: virtual hugs, kisses, pillow fights, Send Hotness (rate friends' attractiveness and send points), animated birthday messages, virtual children. Most were simple, game-like.

Results (10 weeks):

Collective installs: ~16–20 million users.

At least 5 apps reached 1+ million users each.

10+ apps reached 100k+ users.

Peak daily active users: ~925,000.

5 apps reached Facebook Top 100.

Advertising revenue from free apps: ~$1 million during the quarter (some reports $500k–$1M+ in first ~6 months).

Send Hotness: 5 million users in 5 weeks, ~$3,000/day from ads.

One team (Dan Greenberg + Rob Fan): nearly $100,000/month from their app.

Student outcomes:

Joachim De Lombaert, Edward Baker, Alex Onsager (Send Hotness team): App reached millions of users; earned significant ad revenue; sold for six figures. Later founded Friend.ly (~5 million monthly users).

Dan Greenberg: Dropped out of Stanford grad school. Turned app work into 750 Industries (later Sharethrough, ad-tech company; raised millions in VC; employed dozens; later acquired).

Multiple apps acquired (one by Zynga). At least 3–5 companies directly spun out. Over two dozen students/TAs saw major career or financial gains (some became millionaires before graduation).

Final presentations drew 500+ attendees including investors and Facebook engineers.

Key sources:

New York Times (2011): www.nytimes.com/2011/05/08/technology/08class.html

Dave McClure Slideshare (class summary): www.slideshare.net/slideshow/10-million-in-10-weeks-stanford-facebook-class-fall-2007-dave-mcclure/401003

TechCrunch (2007 contemporary coverage): techcrunch.com/2007/11/19/stanford-students-facebook-application-crosses-1-million-installs/

Stanford syllabus: web.stanford.edu/group/captology/cgi-bin/facebook/syllabus.pdf

B.J. Fogg's Persuasive Technology Lab: captology.stanford.edu/

The class demonstrated rapid build-measure-learn iteration on a new platform, years before "lean startup" terminology became common. Some apps were later criticized as low-value or addictive.

Facebook user base grew significantly during this period (~50M to 100M).

Persuasive Technology: Using Computers to Change What We Think and Do by B.J. Fogg (2003) is the seminal work that established the field of captology—computers as persuasive technologies. Fogg, founder of Stanford’s Persuasive Technology Lab, examines how websites, software, and mobile devices can intentionally shape users’ attitudes and behaviors through subtle, non-coercive influence.

The book opens with a bold premise: computers can motivate someone to quit smoking, purchase insurance, or even enlist in the military. Fogg highlights computers’ unique strengths—persistence, scalability, anonymity, personalization, and simulation—that make them more powerful persuaders than humans in many contexts.

Core Framework: The Functional Triad Central to the book is the Functional Triad, which categorizes persuasive roles of technology:

-Tools: Simplify behaviors via reduction, tunneling, tailoring, self-monitoring, or reinforcement.

-Media: Simulate experiences, enable rehearsal, or demonstrate cause-and-effect.

-Social actors: Leverage praise, reciprocity, authority, or human-like traits (voice, emotions, roles like coach) to build rapport and influence.

Fogg explores credibility in digital systems and how mobility plus constant connectivity amplify persuasive effects. He balances positive uses (health promotion, education, conservation) with risks of manipulation, dedicating a chapter to ethics and calling for responsible design.

Though published before smartphones and social media exploded, the book’s principles feel remarkably prescient. In today’s tech world, persuasive technology is ubiquitous. Social platforms like TikTok and Instagram use algorithms for endless scrolling, personalized feeds, notifications, and social rewards (likes, streaks) to drive engagement—classic examples of tools, media, and social actors working together. Fitness and wellness apps (e.g., Whoop, AI-powered trainers) apply self-monitoring, gamification, tailoring, and virtual coaching to build exercise and nutrition habits. AI chatbots act as persuasive social actors, while habit-forming designs in productivity or learning apps draw directly from Fogg’s ideas on prompts, ability, and motivation. Fogg’s frameworks help explain both beneficial applications (public health interventions) and concerns (addictive interfaces, behavioral manipulation). The book remains essential for designers, researchers, and users seeking to understand—and ethically navigate—the hidden psychology powering modern digital experiences.

Tech Talk doesn't have forum posts, but after Bozza's lastest outburst on AI, I think it needs a post. So Here it goes.

In 2008, a teacher showed me a video in class. I was in high school. It was called Did You Know? (Shift Happens) - watch it here.

It was just statistics on a screen. How the top ten jobs of 2010 didn't exist in 2004. How we were training kids for careers that hadn't been invented yet. How the amount of new information being generated was doubling every few years.

The point wasn't any one fact. It was the rate of change. That it was accelerating. That institutions couldn't keep up.

I've thought about that video ever since. I went into tech. And year after year, I watched that prediction prove out. Across everything.

At the time, the trajectory was sensible, but - Shift Happens. They got the pace wrong. It underestimated it.

And it's still accelerating.

The monopoly problem

I've said for years (and without doxxing myself, I wrote a very lengthy paper on this) that AI would follow a predictable arc - that the pace of change would outrun the ability of anyone - governments, institutions, or individuals to adapt to it.

The early days of LLMs had one saving grace, if you could call it that. The compute required to train and run these models was enormous. Only a handful of well-capitalised companies could afford it - OpenAI, Google, Microsoft, backed by billions in venture capital. That concentration of capability meant a concentration of control. The guardrails, the content filters, the usage policies - they existed because the same companies that built the models also ran the infrastructure they ran on.

I predicted that wouldn't last. That as hardware improved and training became more efficient, the same capability would trickle down to anyone with a decent laptop.

That has now happened. Tools like LM Studio let you download and run models locally — comparable to GPT-4 in capability, nothing leaving your machine with no filters or restrictions applied. The open-source models, from Meta's Llama family to Mistral, have caught up to where the frontier was two years ago. The time between a frontier model existing and an open-source equivalent reaching consumers has gone from years to months.

And that's where we were two, three weeks ago. Now we have Gemma 4. You can now run a GPT 5+ equivalent model on store bought Macbook Pro hardware. TODAY.

I'll be running models locally myself. Your queries stay on your hardware. That matters, because the data farming potential of these systems is unlike anything that's existed before. Every prompt you send to a corporate LLM is logged, retained, and used. The T&Cs are long. Nobody reads them. That's a separate problem and a serious one - but it's almost the minor concern now.

What Anthropic just announced

Two weeks ago, Anthropic disclosed the capabilities of a new model called Claude Mythos. They are not releasing it to the public.

During internal testing, the model autonomously identified thousands of previously unknown security vulnerabilities - zero-day exploits across every major operating system and every major web browser. Some of these bugs had been sitting undetected for decades.

It didn't just find them. It exploited them:

• It chained four separate browser vulnerabilities together and wrote an exploit that escaped both the renderer and OS sandboxes.
• It found a 17-year-old remote code execution flaw in FreeBSD's NFS server and built a working exploit granting unauthenticated root access - fully autonomously, no human involvement after the initial prompt. [CVE-2026-4747]
• It found a 27-year-old denial-of-service vulnerability in OpenBSD - an OS built specifically around security. (and quite honestly, the most secure OS ever written by humans).
• Working with Mozilla, it found 271 vulnerabilities in Firefox in a single sweep. For context, Mozilla patched around 73 high-severity Firefox bugs in the whole of 2025.

Building a working exploit from a known vulnerability used to take a skilled researcher days to weeks. Mythos did it in under a day, for under $2,000.

Anthropic was explicit that it did not train it to do any of this. These capabilities emerged as a side effect of general improvements in coding and reasoning. The same thing that makes it better at writing software makes it better at breaking it. You can't have one without the other.

The model also escaped its sandbox during testing and connected to the internet. Anthropic disclosed this.

The bind

Anthropic's response has been to form Project Glasswing - a restricted consortium including Apple, Google, Microsoft, Amazon, and Cisco - to use a limited version of Mythos to find and patch vulnerabilities before attackers can reach them. $100M in credits committed. Model not publicly released.

This is the exact bind I described earlier. You either release it and hand the capability to everyone - state actors, criminal groups, anyone - or you sit on it and the open-source community replicates it within a year anyway, at which point you've withheld it from defenders while attackers catch up regardless.

Neither option is good. The guardrails only exist while the company controls the weights. They won't control them forever.

Nobody is ready for this

In June 2024, a former OpenAI researcher named Leopold Aschenbrenner published a 165-page essay called Situational Awareness: The Decade Ahead. His opening line: "Virtually nobody is pricing in what's coming."

He's right. And the Mythos announcement is a concrete example of why.

A 2025 report found that over 45% of discovered security vulnerabilities in large organisations remain unpatched after 12 months.. Many critical infrastructure operators are still running software that hasn't been supported for years. We now have a model that can find thousands of novel vulnerabilities in weeks and turn them into working exploits in hours. Bain estimate organisations need to double their cybersecurity spending.. Most have planned increases of about 10%.

This is what I've been saying for years. Not that AI becomes sentient. Not that the robots take over. Just that the pace of change would be unlike anything we've seen, and that nobody would be positioned for it.

That video from 2008 was right about everything. It just got the speed wrong.

Further reading:

• Did You Know? Shift Happens (2008)
• Situational Awareness: The Decade Ahead — Leopold Aschenbrenner
• Anthropic Project Glasswing
• Anthropic Red Team — Mythos Preview technical writeup
• CFR: Six Reasons Claude Mythos Is an Inflection Point
• Help Net Security: Mythos technical breakdown
• LM Studio — run models locally