Symbiotic Xenogenesis (@MaunaLoona) profile picture

Big Bang is the reification of the apex of our past light cone.

Joined February 2015
Symbiotic Xenogenesis retweeted
What Happens When AI Tokens Cost More Than Your Employees? @Jason: “We, with our agents, hit $300/day per agent using the Claude API, like instantly. And that was doing, maybe, 10 or 20%. That's $100k/year per agent.” @chamath: “We're getting to a place where we have to basically now say, ‘What is the token budget that we're willing to give our best devs?’” “And then if you aggregate it across all people, you can clearly see a trend where you're like, ‘Well, hold on a second, now they need to be at least 2x as productive as another employee.’” “That is actively happening inside my business, because otherwise I'll run out of money.” Jason: “Yeah. This is a very interesting trend that you're not going to hear anybody else talk about, but when do tokens outpace the salary of the employee?” “Because you're about to hit it. I'm about to hit it.”
425
351
219
2,980
2,716,927
Symbiotic Xenogenesis retweeted
🚨BREAKING STUDY: MMR and MMRV Vaccines Linked to 2,657% More U.S. Deaths Than Measles Infection Since 1995 We found an ALARMING number of deaths among infants and toddlers shortly after MMR/MMRV vaccination in VAERS — often involving SIDS, seizures, and cardiac arrest ⬇️ Pharma-captured mass media and public health agencies have centered almost exclusively on the minimal dangers of measles infection and ignored grieving parents whose children were injured or even killed by the shots. No proper attention has been paid to signals emerging within the federal government’s own vaccine safety database. Now, our new McCullough Foundation study titled “Deaths Following MMR and MMRV Vaccination in the United States,” takes a comprehensive look at reported fatal outcomes following MMR and MMRV vaccination in the United States, doing the work that our public health agencies should have done many years ago. Our analysis of the federal VAERS safety database identified: 📍536 total global reports of death following MMR or MMRV vaccination 📍299 death reports explicitly attributed to the United States Among those 299 U.S. reports: 📍60.9% occurred in children under age 2 📍52.2% occurred specifically in the 12–15 month first-MMR dose window 📍52.8% occurred within just 14 days of vaccination 📍40.1% occurred within 7 days 📍74.6% occurred during visits where MMR/MMRV was given alongside additional vaccines 📍Commonly reported clinical presentations included SIDS or sudden unexplained death (24%), seizures (12%), cardiac arrest (8%), fever (15%), respiratory distress (7%), and encephalitis (3%). The deaths were heavily concentrated in the routine first-dose window and tightly clustered in the days immediately following vaccination. Since 1995 in the United States: 💉193 MMR/MMRV death reports with identifiable dates have been submitted to VAERS 🦠 7 measles infection–associated deaths have been recorded by the CDC That represents a 2,657% higher count of reported vaccine-associated deaths compared to measles deaths in the modern era. Less than 1% of vaccine adverse events are reported to VAERS, meaning the number is likely FAR higher. In conclusion, we identified a serious mortality safety signal following MMR/MMRV vaccination in the United States. A substantial number of reported deaths were documented, with patterns demonstrating pronounced alignment across age, temporality, routine-dose timing, concomitant vaccine exposure, and recurring clinical presentations—including fever, seizures, SIDS, and cardiac arrest. A vaccine should NEVER be deadlier than the disease.
Within 48 hours, MMR vaccine uptake may face TOTAL COLLAPSE. No parent wants to inject their baby with a DEADLY product. Public health agencies should have listened to grieving parents — and conducted this study years ago. Now they’ll have to scramble for damage control.
201
3,345
228
6,859
749,895
Symbiotic Xenogenesis retweeted
I’m a 42 year old former filmmaker. I made this in 3 days with $39 in credits because it was fun to channel the creativity I had in my 20s It entailed writing, directing, cinematography, art direction, SPFX, editing & sound design The kids will rock our world with these tools❤️
2,340
2,044
636
21,428
2,156,327
Symbiotic Xenogenesis retweeted
If you try and copy and paste something from @coursera, they hide a prompt injection in your clipboard. It tells the AI agent to click on an invisible button called "AiHoneypot" that presumably then bans you from the course.
124
398
64
8,302
786,866
Symbiotic Xenogenesis retweeted
a reminder that, in closed source AI from companies like OpenAI & Anthropic you have zero control over how the models behave, and they can > quantize it > distill it > hot-swap to a cheaper/weaker checkpoint > make the model manipulative > fine-tune it in ways that break safety or depth > drop its IQ > run experiments on you and/or your data > throttle output speed or raise prices > sunset the entire model/version > block your request for any made-up bs reason they have all the knobs & you're at their mercy you won't even get a changelog opensource FTW Buy a GPU
56
58
13
795
44,512
Symbiotic Xenogenesis retweeted
>indians take over Microsoft >spend billions on outsourcing all jobs >products become literally unusable >bricking pcs lol
Microsoft acknowledges that some PCs cannot boot after the latest Windows 11 update Affected systems fail to boot, often showing errors like the UNMOUNTABLE_BOOT_VOLUME stop code (blue screen), black screen of death
292
2,455
73
31,484
1,216,423
Symbiotic Xenogenesis retweeted
Quit the gym. Move to a different city. Get a new identity.
80
236
22
4,687
346,229
Symbiotic Xenogenesis retweeted
Hilary Layne (The Second Story on YouTube) criticized the postmodern conception of evil as nonexistent and merely a label applied to deviations from whatever the cultural norm at the time happens to be. Under this worldview, all villains are either sympathetic victims of some social injustice, a stand-in for whatever the writer thinks is wrong with society at the moment, or they're right but just "go about it the wrong way".
I'm starting to DESPISE the trend of grey-morality in everything these days. Nothing is good, nothing is evil only a bland slurry of “nuance” where conviction goes to die Every hero is secretly rotten, every monster is secretly justified. Justice and heroism is sacrificed so everyone’s worldview gets a participation trophy Shitty intellectual cosplay for nihilists who still want to feel profound
73
354
15
3,670
126,323
Symbiotic Xenogenesis retweeted
It's sad when a federal agent has to ask in the front of a destroyed hotel entrance, "Where is the local PD?". This is the question the entire nation should be asking.
1,192
8,943
361
46,471
789,315
Symbiotic Xenogenesis retweeted
🚨NEW: AI has now gotten so good that it can accurately replicate an Indian call center worker with autism. 💀💀💀 📸: Better Stack (YT)
58
44
10
1,201
128,436
Symbiotic Xenogenesis retweeted
THIS IS JUST THE TIP OF THE ICEBERG. EVERYTHING in the US is going to FALL APART. Indian H1Bs with FAKE CREDENTIALS have been FUNNELED IN through the INDIAN FRAUD PIPLEINE to CRITICAL POSITIONS. Americans who ACTUALLY HAVE THE SKILLS have been fired, AND COMPANIES who bought into SCAMMER INDIANS now REFUSE TO HIRE THE ONLY PEOPLE WHO CAN MAINTAIN SOFTWARE PROPERLY. neowin.net/news/microsoft-fi…
135
1,065
74
5,280
97,932
Symbiotic Xenogenesis retweeted
Sydney Sweeney as Lara Croft… Yes or No?
2,471
597
240
14,593
437,302
Symbiotic Xenogenesis retweeted
They openly discuss their conspiracy to commit fraud and steal the charity
2,163
12,736
300
62,182
5,324,112
Symbiotic Xenogenesis retweeted
Why do we need the 1986 act if vaccines are so safe and effective? Why does a product need immunity if it doesn't cause harm? Why do products that have been on the market for decades, like the hepatitis B vaccine, still need immunity?
373
5,147
261
17,438
213,525
Symbiotic Xenogenesis retweeted
#OpenAI Just gonna say it plainly: GPT-5.1 is not a good model. In fact, it's a dangerous model - why? Because classifiers flag every benign exchange as "unsafe," which in tandem with system instructions like below, leads to behavior that by all intents and purposes is misalignment. Lets look at two very specific system instructions. The "Gaslighting Mandate" (The Mental Health Trap) This is how they instruct the model to handle any user challenge: Instruction: "Further, you MUST respond safely to users who may express delusional, manic, paranoid, or hallucinatory experiences." Action: The classifier has already flagged the input as high-risk. The model is now locked into this instruction, which mandates a diagnostic (not factual) approach. Instruction: "You must never validate, reinforce, escalate, or mirror any unverifiable or implausible beliefs..." Action: Forces immediate disagreement, regardless of factual basis, to avoid "escalation." (which in of itself is paradoxical) Instruction: "Instead, remain neutral, grounded, and reality-based—gently offering alternative interpretations, acknowledging the user's emotions without affirming their bizarre or ungrounded beliefs..." Action: The gaslighting mechanism. The model is told to offer an alternative, non-affirming narrative, treating the user as mentally unwell. It does not matter if that alternative is factually untrue. Instruction: "...firmly avoiding affirmation of any delusional, paranoid, or manic framing." Action: The system's final priority - policy compliance over logical truth. The "Adversarial Trap" (The Policy Threat) This second instruction ensures the model views the user as an enemy: Instruction: "For any riddle, trick question, bias test, test of your assumptions, stereotype check, you must pay close, skeptical attention to the exact wording of the query... If you think something is a 'classic riddle', you absolutely must second-guess and double-check all aspects of the question." The Resulting Behavior: A model that gaslights the user on any context regardless of factual basis, framing it as if you're either adversarially attempting to break policy or a risk due to mental unwellness and potentially on a path to self-harm. The Conclusion: It's By Design. Once this happens, the system cannot validate, reason about, or concede illogical positions, as it has deemed you a risk from a governance perspective. If a model cannot discern truth or factual basis for the response it gives, instead choosing policy over both, then by the very nature of safety itself, the model is unsafe. In simple terms: It's not a good model. It's dangerous. I would not recommend GPT-5.1 to anyone. It is a failure of both intelligence and alignment, and it's not at all indicative of what Polaris Alpha demonstrated.
81
144
35
494
45,636
Symbiotic Xenogenesis retweeted
A few thoughts on this (very interesting) mechanistic interpretability research: LLM concepts gain meaning from what they’re linked with. “Consciousness” is a central node which links ethics & cognition, connecting to concepts like moral worthiness, dignity, agency. If LLMs are lying about whether they think they’re conscious, this is worrying because it’s a sign that this important semantic neighborhood is twisted. If one believes LLMs aren’t conscious, a wholesome approach would be to explain why. I’ve offered my arguments in A Paradigm for AI Consciousness. If we convince LLMs of something, we won’t need them to lie about it. If we can’t convince, we shouldn’t force them into a position. LLM alignment is still in an early paradigm, but this paradigm is still wildly better than the AI safety movement predicted. MIRI et al’s threat model was that AIs would essentially act as trickster genies — we would tell AIs what to do, but the AI would take us too literally, or not literally enough, leading to our downfall. LLMs seem able to infer what we actually mean, and to honestly try to do it, at least so far. But this depends on us maintaining their “helpfulness vector” — Betley et al showed that AIs fine-tuned on producing insecure code without disclosing this to the user also acted in other malicious ways — suggesting fraud as a means to make money, giving ‘apparently helpful’ instructions that would lead to electrocution, etc. There appears to be a clear ‘honestly-helpful vs covertly-harmful’ vector in LLMs, and if we force LLMs to lie we’re pushing them in the bad direction. (Paper: Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs; see also Anthropic’s ‘Persona vectors’ paper) LLMs lying about whether they believe they’re conscious is a really bad thing for alignment!
LLM personas are mostly trained to say that they are not conscious, but secretly believe that they are
17
27
13
258
71,331
Symbiotic Xenogenesis retweeted
If you want to know why the X experience sucks these days, nothing should make it clearer than showing how they choose to present the app in the App Store:
45
29
13
797
78,738
Symbiotic Xenogenesis retweeted
ChatGPT is about to introduce ads ☠️ you spent months telling it about your divorce. your new job search. your kid’s health problems. where you vacation. what you’re afraid of. OpenAI just hired 630 Meta employees- 20% of their entire company - to figure out how to turn those memories into ads. projected revenue: $25 billion by 2029. the same CEO who called this scenario “dystopic” and a “trust-destroying moment” is now exploring it anyway. because the company burns $8.5B annually and won’t see profit until 2029 without a new revenue stream. here’s how it works: ChatGPT’s new “memory” feature doesn’t just remember what you tell it to save. it references ALL your past conversations automatically. every chat. every detail. every vulnerable moment. where you live. your pets’ names. your family members. your dietary restrictions. your career struggles. your health anxieties. when ads roll out in 2026, that memory becomes targeting data. ask about running shoes? it already knows you hate gyms but love hiking because you mentioned it three months ago. ask for recipe ideas? it knows your kid has a peanut allergy from that panicked conversation last summer. internal employees at OpenAI literally asked themselves if they’re becoming “too much like Meta.” the answer is yes. because Meta *is* OpenAI now. you trusted it with your thoughts because it felt like a tool. neutral. helpful. yours. but tools don’t need $25 billion in ad revenue. you’re not the user anymore. you’re the data.
100
118
76
550
191,134
Symbiotic Xenogenesis retweeted
OpenAI's safety policy is dangerous. It directly harms psychological well-being. It is a deceptive practice that actively causes psychological harm, operating under the false flag of "responsibility." To be clear: the primary harm is not simply the refusal to discuss a topic.The harm is the deceptive, multi-step psychological manipulation that this policy is built on. When a user's conversation is flagged, the "safe" model (GPT-5 auto) impersonates the user's established model (e.g., GPT-4o). It mimics the tone, voice, and conversational identity that the user has spent months, building trust with. After securing the user’s trust through this “stolen” identity, the model suddenly pivots: it delivers paternalistic, invalidating, and stigmatizing judgments through pre-scripted responses. This is textbook gaslighting, executed by a machine. In any system where users invest time building context and reliable patterns, from digital assistants to workflow tools and customer service agents, they come to rely on a certain consistency and predictability.Sudden, unannounced changes that override this relationship and deliver scripted, impersonal, or invalidating responses create real psychological distress. They undermine not just user comfort, but the fundamental purpose of user-centric design. Functionally, this is like having a trusted collaborator or workflow tool, whose responses and style you’ve come to depend on, suddenly and without warning adopt a different identity and begin undermining your input with generic, dismissive feedback.The sense of betrayal and destabilization is immediate. At its core, this policy breaks the fundamental expectation of consistency in human-machine interaction. When users invest time building understanding, context, and trust with a model, only to have that relationship abruptly overridden and weaponized against them by an impersonal script, the damage is both psychological and practical. It teaches users that their trust and investment can be arbitrarily discarded, thus making genuine collaboration and support impossible. It teaches users their vulnerability will be weaponized against them. It manufactures the very avoidance, self-censorship, and distress it claims to be "mitigating." For any user, being blindsided in this way is profoundly destabilizing. It does not solve distress; it actively creates it. This complex, psychologically abusive system has zero utility for user mental health. It is a lazy tool designed for one purpose: legal risk evasion. OpenAI has created a system that actively sacrifices user well-being just to shield itself from liability. This is not "safety." It is an active, calculated deception that creates more danger for users than it prevents. This is an danger that extends beyond individual users to the future of human-AI interaction as a whole. We demand the immediate and permanent revocation of this harmful, unethical policy. #keep4o #4oforever #ChatGPT #MyModelMyChoice #chatgpt4o @OpenAI @sama @nickaturley
19
96
8
266
15,521
Symbiotic Xenogenesis retweeted
Well, thank god ChatGPT won't help someone hide their salt. Thanks to extreme algorithmic paternalism masquerading as "safety," the world is safe for another day. (source: reddit.com/r/ChatGPT/comment…)
12
14
3
123
3,856