i think it boils down to: people are stupid, lazy, or ignorant
im not taking a stance. but that would be my shot at the root cause.
those may sound like negative dehumanizing words. but those are very natural,human, vices.

tennessee, florida, georgia . if you live in a crappy place… you know it. move somewhere better.
i see the concern. but takeaway could be percieved as: “cops using cameras as cameras”.
whats the issue?

it would be a better fit for intellectualproperty community . or maybe security. i guess it sorta could be privacy. just not the way most people tend to focus on. intellectualproperty is the bigger issue. but it occured due to lack of privacy .

this is the whole selling-point of locallama and why local models are so rigorously sought after (despite seeming excessive(seemingly needless) spending on rigs )

opensource Lean specialized models are available to anyone: https://huggingface.co/mistralai/Leanstral-1.5-119B-A6B

i think in some ways this is a good thing. for example you can rent low resource compute for as low as $0.10/month . its a global fiercely competitive market these days.

i feel the pain.
yet i am cautiously optimistic .

homeless ( zero-rent )

RE: xiaomi 18 fold :

• likely initially china only
• actual ram type is unknown

LLM answer:

- Tracking Pixels and Web Beacons
- Email-Open Tracking
- Email Link-Click Tracking
- Advertising Impression Tracking
- Conversion Tracking
- Social-Media Pixels
- Analytics Tags and Tag Managers

- IP Address and Network Tracking
- IP-Based Geolocation
- ISP, ASN, and Organization Identification
- Household and Network-Level Matching
- VPN, Proxy, and Tor Detection
- Network Reputation and Fraud Scoring

- URL and Referral Tracking
- UTM Parameter Tracking
- Advertising Click-ID Tracking
- Affiliate and Referral-ID Tracking
- Email Campaign-ID Tracking
- Link Decoration
- Redirect-Chain Tracking
- Cross-Site Identifier Propagation
- HTTP Referer Header Tracking

- HTTP Header and Cache Tracking
- User-Agent Tracking
- Client Hints Tracking
- Accept-Language Tracking
- ETag Tracking
- Cache-Control and Last-Modified Tracking

- Browser Storage Tracking
- First-Party Cookies
- Third-Party Cookies
- LocalStorage and SessionStorage Tracking
- IndexedDB Tracking
- Cache Storage Tracking
- Shared Storage Tracking
- Service-Worker Storage Tracking
- Evercookies and Identifier Respawning
- Supercookies
- HSTS Supercookies
- Favicon Cache Tracking
- TLS Session-Resumption Tracking

- Browser Fingerprinting
- Browser, Version, and Operating-System Fingerprinting
- Screen, Viewport, Color-Depth, and Device-Pixel-Ratio Fingerprinting
- Time-Zone, Language, Locale, and Keyboard-Layout Fingerprinting
- Installed-Font and Font-Rendering Fingerprinting
- Canvas Fingerprinting
- WebGL and WebGPU Fingerprinting
- AudioContext Fingerprinting
- GPU and Graphics-Driver Fingerprinting
- CPU Core, Device-Memory, and Hardware-Concurrency Fingerprinting
- Touchscreen, Input, and Device-Capability Fingerprinting
- Camera, Microphone, and Media-Device Fingerprinting
- Browser Extension, Plugin, MIME-Type, and PDF-Viewer Fingerprinting
- CSS Feature and Media-Query Fingerprinting
- Browser API Availability Fingerprinting
- WebRTC and Local-IP Exposure
- Network Information API Fingerprinting
- TLS, JA3, and JA4 Fingerprinting
- HTTP/2 and HTTP/3 Protocol Fingerprinting
- DNS Resolver Fingerprinting
- Clock-Skew and Performance-Timing Fingerprinting
- Sensor Fingerprinting
- Motion, Orientation, and Ambient-Light Sensor Fingerprinting
- Speech-Synthesis Voice Fingerprinting
- Emoji and Text-Rendering Fingerprinting
- Math, Floating-Point, Error, and Exception Fingerprinting
- Browser Automation and Bot-Detection Fingerprinting
- Combined and Probabilistic Fingerprinting

- Account and Login-State Tracking
- Logged-In Account Tracking
- Login-State Detection
- Single Sign-On Tracking
- OAuth and Social-Login Tracking
- Embedded Social-Widget Tracking
- Account-Recovery Identifier Matching

- Cross-Device and Identity-Graph Tracking
- Login-Based Device Linking
- Hashed Email and Phone-Number Matching
- Probabilistic Device Matching
- Data-Broker Identity Resolution
- Offline-to-Online Data Matching
- CRM and Customer-Data Matching

- Server-Side Tracking
- Server-Side Analytics
- Server-Side Tag Management
- Server-to-Server Event Tracking
- Advertising Conversion APIs
- Purchase, Order, and Lead-Event Sharing

- Behavioral Analytics and Session Replay
- Mouse-Movement Tracking
- Scroll-Depth Tracking
- Click and Hover Tracking
- Keystroke and Form-Interaction Tracking
- Heatmap Tracking
- Rage-Click Detection
- Dwell-Time and Engagement Tracking
- Form-Abandonment Tracking

- Push-Notification Tracking
- Web Push Subscription Identifiers
- Mobile Push Tokens
- Notification Engagement Tracking

- Mobile-App Tracking
- Mobile Advertising IDs
- App Instance and SDK Identifiers
- App-Install Attribution
- Deep-Link Tracking
- Mobile Device Fingerprinting
- Mobile Location Tracking

- Location Tracking
- GPS Location
- Wi-Fi Network Location
- Bluetooth Beacon Location
- Cell-Tower Location
- Nearby Device and Network Discovery

- Embedded Third-Party Content Tracking
- Embedded Video Tracking
- Embedded Map Tracking
- Comment-Widget Tracking
- Payment-Widget Tracking
- CAPTCHA and Anti-Bot Widget Tracking
- Third-Party JavaScript Tracking

- DNS and Domain Cloaking
- CNAME Cloaking
- First-Party Subdomain Tracking
- Domain Alias Tracking
- DNS Query Tracking
- DNS Prefetch Tracking

- Advertising and Attribution Tracking
- Real-Time Bidding Tracking
- Ad-Exchange and Demand-Side Platform Identifiers
- Retargeting and Remarketing
- Frequency-Capping Identifiers
- View-Through Attribution
- Multi-Touch Attribution

i think intelligence agencies just collect data on everyone and for any profitable purpose. so i dont exactly** think that this is a new threat. its the same threat but a little more focused.

hell yeah

12G ram and 2TB extendable storage is very sufficient among most of competitors.

i dont want to share it. but it was easy to add a little more description for basic criteria in the request. although, i did not detail a step-by-step process for drawing.
i had this idea to feed it an actual image, to base the drawing on. but i wasnt able to. llama.cpp seems to be messed up[esp vulkan pkg](and also i probably dont know the proper way to utilize features(if they even exist for a model))

i was using qwen and somehow it offered a how-to guide about how to gently help a loved-one escape a cult. it seemed random but, the more i think about it… it seems applicable to lots of life problems. idk what it says about society or humans. maybe just that stubborness and fixed-thinking is a common, difficult, and harmful problem in society.

im not bothered by it.
but also… i wouldn’t use it as my sole model to perform every task. [waves hand toward huggingface.] ( a gazillion bajillion models, specialized for all types of tasks).

edit: i checked. and there aren’t as many models as i expected, for this task(didnt test them). and… small multimodal model failed badly(see picture of the iguana … or whatever the hell that is)
it’s a strange benchmark. but it seems reasonable to expect LLMs to be able to do this.

gemini://kennedy.gemi.dev/

evidence of preparation on modelscope changelogs.
also “more sizes”

q 35b a3 is expected soon!

theres also https://europepmc.org/