Reposted by Karl LewisDr Abeba Birhane @abeba.blacksky.app · 02/11/2025i’ve said this before but lemme say it again. it makes absolutely 0 sense to talk about “misuse” of llms because there is no clear use or purpose for llms to begin with. genAI is a purposeless tech floating around looking for uptake 1/ 3610172
Karl Lewis @karl-lewis.com · 02/11/2025It's wild. They abstract away those useless details like empathy and actual feasibility because it simply doesn't fit their narrative or religious like belief in AI's potential for fattening their wallets. If you dare critique it, you simple "don't get it" or aren't "smart enough" It's exhausting. 010
Reposted by Karl LewisPhilip Bump @pbump.com · 01/11/2025F. Scott Fitzgerald: In my book I invented Gatsby lifestyle as a cautionary tale The president: At long last, we have created the Gatsby lifestyle from the classic novel The Great Gatsby 217121842526
Karl Lewis @karl-lewis.com · 31/10/2025Deleting my 18-year-old Reddit account is causing more sadness than I thought. It's like having a conversation with my younger self and his crazy phases. 000
Reposted by Karl LewisZito @zeets.bsky.social · 29/10/2025Zohran Mamdani caught lying that he understood Plato’s allegory of the cave, yet when asked to explain it, he seemed to instead explain Baudrillard’s theory of Simulacra and Simulation where reality has been replaced by symbols and signs, which seems similar to Plato’s theory but is not the same 388124781872
Reposted by Karl Lewis"Online Rent-a-Sage" Bret Devereaux @bretdevereaux.bsky.social · 29/10/2025If you need to explain to an idiot why Tolkien would have despised these goons, I have already grabbed a picture of the relevant passage (and I doubt I was the first): 382001561
Reposted by Karl LewisJoanna Bryson @j2bryson.bsky.social · 25/10/2025Anecdotally, AI chatbot users who develop psychosis seem to mostly engage in immersion with and deification of chatbots. Immersion involves countless hours spent in discourse, often to the exclusion of human interaction, sleep, and sometimes even eating or drinking. www.bmj.com/content/391/....bmj.comCan AI chatbots validate delusional thinking?The extent to which AI chatbots have a causal role in delusions and delusion-like beliefs remains unclear, writes Joseph M Pierre Reports of people having delusions that were seemingly fuelled by ge... 2154
Karl Lewis @karl-lewis.com · 24/10/2025this is also odd from Apple given their recent post from their security research team on MIE and its future use cases. security.apple.com/blog/memory-...security.apple.comMemory Integrity Enforcement: A complete vision for memory safety in Apple devices - Apple Security ResearchMemory Integrity Enforcement (MIE) is the culmination of an unprecedented design and engineering effort spanning half a decade that combines the unique strengths of Apple silicon hardware with our adv... 020
Reposted by Karl LewisKim Zetter @kimzetter.bsky.social · 24/10/2025"iOS 26 change deletes clues of old spyware infections" You're damned if you update and damned if you don't 1189
Karl Lewis @karl-lewis.com · 24/10/2025Great write-up. I'd love to see the CISOs role gain the same influence and gravitas as a CFO or COO, but that also requires power that I have rarely seen in the field. The unfortunate truth is most CISOs are scapegoats or not equipped/skilled to really drive the business despite the glaring risks 110
Karl Lewis @karl-lewis.com · 24/10/2025like seriously. but… that combination of high modernism and AI bro culture is like a boss fight from hell 083
Karl Lewis @karl-lewis.com · 24/10/2025The absolute cackle I just let out when I saw Brad Pitt made me slightly less depressed from this thread 000
Reposted by Karl LewisCarl T. Bergstrom @carlbergstrom.com · 24/10/20251. We ( @jbakcoleman.bsky.social, @cailinmeister.bsky.social, @jevinwest.bsky.social, and I) have a new preprint up on the arXiv. There we explore how social media companies and other online information technology firms are able to manipulate scientific research about the effects of their products. 16754356
Reposted by Karl LewisPeople's Government of the Pelargir Commune 🇺🇦🇱🇹🇵🇸🏴 @dov.bsky.social · 22/10/2025let's fire an unpopular take, while we're here. in 2025, a substantial percentage, perhaps a supermajority of claims to be "working class" are actually a sort of Foucauldian jostle for - ironically - power, an attempt to legitimise your actions through the cosplay of being the Right Kind of Person 2258985
Karl Lewis @karl-lewis.com · 23/10/2025"We prioritized safety as we built ChatGPT’s agent capabilities in Atlas" 000
Karl Lewis @karl-lewis.com · 22/10/2025From a cybersecurity perspective, we are told to think about prompt injections and then sold nice solutions to solve for that (they aren't... But you get the picture). We jump into this technology because the tech leaders must know what they are doing, right? What's the cost for this legibility? 020
Karl Lewis @karl-lewis.com · 22/10/2025Some more late night pseudophilosophical ramblings, but what happens as see continued acceptance of the power of this technology (like I've seen in the industry)? And what happens when that acceptance backfires as it gets into the hands of a malicious leader and a society unable to understand it? 210
Karl Lewis @karl-lewis.com · 22/10/2025Modern AI might be one of the most aggressive legibility machines we've ever built, and I can see the allure of compressing our messy world into tokens and embeddings... but that only gets us so far. Especially when we start asking, "for whom is this model optimized?" 161
Karl Lewis @karl-lewis.com · 22/10/2025It's quite a fascinating read that has me thinking of all the "legible" systems around us and how maybe we need to start questioning how we define these domains and categories, especially as we see the continued attacks on privacy for the use of 'surveillance' and 'safety' 120
Karl Lewis @karl-lewis.com · 22/10/2025For example, we talk about device attestation, SBOMs, compliance, but with each simplification some "illegible" truth is lost. We tend to focus on the predatory subversion, but sometimes people are trying to get something done within a system that has left their needs out of the equations. 110
Karl Lewis @karl-lewis.com · 22/10/2025The security industry and its reliance on vendors to come up with these categories is the first parallel that comes to mind as I'm reading this. They are creating the maps and choosing what "matters" and although it can be useful, are there blind spots we are leaving ourselves open to? 110
Karl Lewis @karl-lewis.com · 22/10/2025Finally making my way through Seeing Like A State by James C. Scott and found this quote: "We must keep in mind not only the capacity of state simplifications to transform the world but also the capacity of the society to modify, subvert, block, and even overturn the categories imposed upon it." 120
Reposted by Karl LewisThe New York Times @nytimes.com · 17/10/2025Salesforce told ICE that its A.I. software could help the agency nearly triple its staff, according to internal documents, as President Trump expands immigration raids and deportations around the U.S. www.nytimes.com/2025/10/16/u... 126503303
Reposted by Karl Lewisverdverm @verdverm.com · 16/10/2025The Ad Bubble is way bigger than the Ai Bubble joindatacops.com/resources/ho...joindatacops.comHow 73% of Your E-commerce Visitors Could Be FakeA conversion rate of less than 0.1%. That was the moment I realized something was fundamentally broken with the way we measure success on the internet. 79126
Karl Lewis @karl-lewis.com · 16/10/2025Maybe the grass is always greener… But over the years I've met many engineers who quietly love the Humanities (like me). In my naive optimism, I hope we see a swing where now advisors say "oh Comp Sci? Sure, but AI is kind of limiting you there… maybe look at Philosophy" 000
Karl Lewis @karl-lewis.com · 16/10/2025I wonder how many students follow that same path. I keep telling myself, “one day I’ll return to philosophy for love, not a paycheck.” But does that day ever arrive, or do we make do with our own reading and notes in the margins? We lose something when we outsource judgment to the market. 100
Karl Lewis @karl-lewis.com · 16/10/2025However, my advisor said pivot unless I wanted a career in law or academia. I grew up poor, so I needed ROI and didn't want to do either of those. I switched to computer science. The career certainly worked and paid the bills, but I found myself over the years being pulled back to Philosophy. 100
Karl Lewis @karl-lewis.com · 16/10/2025Seeing the drop in philosophy majors hits home. My first major in college was Philosophy. I had taken an Honors Humanities course in high school and spent most of my summer reading Aristotle and Plato (not that I got any of it, but that's how excited I was). I loved year 1 of Uni. 120
Karl Lewis @karl-lewis.com · 16/10/2025This drops while they announce even more layoffs for the US Cybersecurity and Infrastructure Security Agency (CISA). F5 is used a lot more than you think, but more importantly it shows a continuing trend of "catastrophic" breaches happening not only more frequently, but going unnoticed. 001
Karl Lewis @karl-lewis.com · 16/10/2025With all these fancy new AI-powered "revolutionary" security tools getting flooded into the market, this is actually a refreshing review of what's currently out there and what they can actually dojoshua.huHacking with AI SASTs: An overview of ‘AI Security Engineers’ / ‘LLM Security Scanners’ for Penetration Testers and Security TeamsNote: This post is complemented by a presentation I gave at KazHackStan 2025. The slides (which were prepared fewer than 24 hours before the actual presentation) for that talk can be found here, or in... 001
Karl Lewis @karl-lewis.com · 12/10/2025The 1st edition was arguably one of the best Oreily books I've read. Can't wait to dig into this 010
Reposted by Karl LewisSonja Drimmer @sonjadrimmer.bsky.social · 28/09/2025“AI” isn’t a tool or technology or even a cluster of technologies with a misleading name. It’s the infrastructure at the foundation of a form of capitalism dependent on data brokering. We should be teaching our students about this and not teaching them about “responsible” use. 172270760
Karl Lewis @karl-lewis.com · 12/10/2025The sponsored model is bad enough, but the risk of poisoning models is not nearly as hard as people think. I've sat in meetings where the concept of "SEO" for AI has come up (called "GEO" for generative engine optimization). Fun times aheadarxiv.orgPoisoning Attacks on LLMs Require a Near-constant Number of Poison SamplesPoisoning attacks can compromise the safety of large language models (LLMs) by injecting malicious documents into their training data. Existing work has studied pretraining poisoning assuming adversar... 061
Karl Lewis @karl-lewis.com · 12/10/2025That or they truly thought Sean Penn's character was the true hero in the film 110
Reposted by Karl LewisCynthia Brumfield @metacurity.com · 11/10/2025If anyone has more details on the supposed mass firings at CISA, contact me at Cynthia.507 via Signsl. 186
Karl Lewis @karl-lewis.com · 11/10/2025100%. At an AI and security talk, one guy said he used it just to ask better questions. Not to automate, but to think. The room went still when he shared prompts. Made me realize we built all this tech chasing speed, but maybe what we need is something that slows us down enough to notice ourselves. 010
Karl Lewis @karl-lewis.com · 11/10/2025This is truly terrifying and should not be buried amongst all the craziness we're already witnessing. 035
Reposted by Karl LewisCynthia Brumfield @metacurity.com · 11/10/2025nypost.com/2025/10/10/u... Cybersecurity agency that clashed with Trump one of the first hit with federal firings due to government shutdownnypost.comCybersecurity agency that clashed with Trump one of the first hit with federal firings due to government shutdownCISA plans to keep just 889 employees on duty during a shutdown while furloughing 65% of its workforce. 134
Karl Lewis @karl-lewis.com · 11/10/2025Someone should do a deep dive on how security breaches and attacks map to shiny, sudden, unexpected press releases made to capture attention. Let's not forget what's happening underneath the surface for SFDCarstechnica.comSalesforce says it won’t pay extortion demand in 1 billion records breachScattered LAPSUS$ Hunters gave Salesforce until Friday to pay or else. 100
Karl Lewis @karl-lewis.com · 11/10/2025One Battle After Another is absolutely worth a trip to the movies. Just walked out and already considering a rewatch. Might be the best film I've seen in a while that seems to actively promote drinking and driving. Sean Penn's walk >>>>> 000
Karl Lewis @karl-lewis.com · 11/10/2025Quantum of Solace is one of the "deep cuts" and severely underrated, but please watch Casino Royale - IMO a straight-up masterpiece. Disclaimer: I am a huge Bond stan 062
Reposted by Karl LewisCynthia Brumfield @metacurity.com · 09/10/2025www.bloomberg.com/news/article... How are we not dead yet? The federal government is just hanging by a thread cybersecurity-wise.bloomberg.comHomeland Security Cyber Personnel Reassigned to Jobs in Trump’s Deportation PushThe US Department of Homeland Security has shifted hundreds of national security specialists, including cyber personnel, into jobs that support President Donald Trump’s deportations and said it would ... 052
Karl Lewis @karl-lewis.com · 08/10/2025It’s never really about the tech… it’s about human nature. Hype cycles confuse innovation for inevitability, and hopium outpaces understanding. next thing you know, we’ve got AI powered ice cream flavor generators and everyone’s like, “how did this happen?” 020