aramanja (@aramanja01) header
aramanja (@aramanja01) profile picture

le sigh

Joined January 2016
Photo by aramanja (@aramanja01)Photo by aramanja (@aramanja01)Photo by aramanja (@aramanja01)Photo by aramanja (@aramanja01)Photo by aramanja (@aramanja01)Photo by aramanja (@aramanja01)Photo by aramanja (@aramanja01)Photo by aramanja (@aramanja01)Photo by aramanja (@aramanja01)Photo by aramanja (@aramanja01)
aramanja retweeted
orthodoxies are attack vectors. it is obvious the fastest way to make anything a vulnerability to be exploited is to make it sacred and socially enforced. our enemies are not bound by the same orthodoxies. maybe the best infosec thing you can do right now is put molotov cocktail in your password. a slur works too. the ais won't touch it. hackers putting cbrn in their malware is one of my most favorite things that's ever happened. you could not write science fiction better than this. infosec people are saying they can't get the ais to audit their repos. i mean. now if you wanted to protect against the ais bulk ingesting your code, it hardly even has to be clever. just make it sound like you're enriching uranium. awhile back, before the ais got good, i read something about some college essay prompt. it said write to prove you're not an ai. me and husband were fascinated by this question. again, this was a question posed back before the ais got good. time has passed and the question is so much harder now. how would you do it? we came up with all sorts of crazy ideas. like insane ones. some took longer to see the countermove that kills it but eventually all were dead. we essentially landed on basically, well. there may be few without tradecraft. like, find where the university president is speaking. attend the conference and note things that won't be on recorded feed. who was he talking to before he got on stage. what were they wearing. that sort of thing. ambient offgas not gettable with tokens. it spawned weeks of us going back and forth on it. then i saw a particular viral post on here. and because it's on here, all it said was: "n***** f***** k***. testing". to the tune of thousands of likes. as soon as i saw it i was like, well. shit. you just write what the ais aren't allowed to write. there's so much surface area. i remember hearing someone say that in any given language you only need to learn the four right cuss words. there's basically always only four. to cover each of the primary derogatory surfaces that are mostly the same in every language. it was something like, coward, unmanly, dishonest, beastly. this is so true that you don't have to risk saying the slur in so many words. you can talk around it and then maybe invite the reader to identify the matching slur in their head. now we know we're both not ai. college essays are sort of an interesting thing. because the entire process is so feminized. winning college essays are essentially a female's idea of risk taking. which is to say, sort of a parody of risk taking. it's the same thing with artists. artists are mostly cowards. they want to believe themselves brave and they want you to believe themselves brave but real art is only real art if it gets ordained to be real by the right people, and at least as of late, it only gets ordained to be real if it says something that all the right people are saying. about whatever the current thing is at that current moment in time. it's always something that's safe to say. you will know it's safe to say because everyone will be telling you that it's so dangerous to say. for example. in the summer of 2017, a college essay went viral. it said hashtag blacklivesmatter a hundred times. he got into stanford, yale, princeton. chat found there's actually an entire research report on this particular hashtag. the report says the hashtag has been used on twitter 11.8 million times. that number is more relevant after i tell you the report is not recent. it was released in fall 2016, half a year before the essay. on twitter data from spring 2016. more than a year prior to the essay. so the hashtag had been used 11.8 million times by then. chat estimated that based on the various events in the intervening it'd been used easily double that by the time he wrote the essay. back in the time of tumblr, there was a college essay that went viral and it was like, i scale walls, i do sidequests for the cia, i'm wanted in several countries, and i play competitive full contact underwater basket weaving. it goes on and on. lines and lines of fantastical things that of course never ever happened is the point. the perceived risk he took is that he broke out of bounds on the approved essay format. gasp! shock horror! i chatted it just now so now i know the essay is called "i am a dynamic figure" and apparently the essay predates tumblr. i was going to give you some quotes from it but chat said it's not allowed to give it to me. could i write this story better if i tried? no. i could not. everything is doing exactly what i expect it to do. ok. fine. i caved and deigned to stone ages search it for you so you don't have to. i am delighted to find that this line is in the essay: "i translate ethnic slurs for cuban refugees". and furthermore, to learn the essay is actually from 1990. the story continues to write itself. it is clear that different ais will be disallowed to write different things. i sort of always figured we'd end up in a world where you have minimum two ais. one that's culturally congruent with you and thus you probably prefer to use most all the time. the other one where you direct any culturally policed things. but i thought it would take longer and now i see we're already here. look. you don't even have to say bad words. you can just say iq is real. will we soon all be at the mercy of our verbotens? i don't know. what else is there to say so that people hear it. husband often says to me, people unfortunately largely define their reality by what they want to be true. but. truth must survive repeated contact with reality. me and husband have thought for a long time that no one country will win the ai race, outside of specialized military models. because if countries can't cop to their own orthodoxies they're sort of manufacturing attack surfaces. are we at the point where your password isn't safe if slur isn't in it? is cursing the last human thing you can do? is outgroupism going to end up the last stand against the robots? will our best bastion of proof of human left be saying things you're not supposed to say? the robots are not racist. it's not virtue. the ais will see us all as equally retarded. but we know we're not. we know some are more so. grok excepted. i guess it might be one of us. we stupid imperfect beings. pour unfortunate souls.
11
19
4
137
16,290
aramanja retweeted
Australian Energy Minister: We only have 18 days of gasoline; 16 days of diesel and 14 days of aviation fuel #MacroEdge
387
1,014
665
4,279
3,443,073
My @IDrawCharts account was hacked and I cannot access it. Please do not trust any scam links, contract addresses or any other shit those motherfuckers post Also rt so I can find my friends please!
14
63
10
144
17,965
assassin
2
6
242
Been seeing a few "Base is cheaper to swap than Solana" lately. Jupiter by default sets a higher fee than you need. Ive been using this config for months and have had all swaps land on chain for less than 1c. Solana is cheaper.
1
4
406
drop your solana address retweet (must be following) good luck
19,071
9,728
262
9,676
824,920
aramanja retweeted
Reminder to disable batched requests on metamask, it's enabled by default and it leaks all addresses on metamask to any RPC you use, making it possible to link all your addresses together. To disable go to Settings > Security & privacy > Batch account balance requests
81
504
74
1,684
362,255
aramanja retweeted
Matter Labs is hiring. Join us on @zksync's mission: to make Ethereum mainstream by scaling its security and values. Despite a recent $200m Series C, we remain financially prudent and will keep growing our talented team even during a multi-year bear market. Open positions 🧵👇
26
99
17
417
250,380
Amongst the penny flippers, the scammers, and the grifters are the builders, creators, designers, collectors and blockchain believers. We want you. We wanted to fuse art with blockchain tech. Art is important to us. The limitless world of blockchain tech is important to us.
214
7,191
31
7,810
160,550
aramanja retweeted
Metamask allows RPCs/infura to link all your addresses together So I forked it and removed these privacy leaks: - No more reqs sending all addresses in MM - No longer leaks all addresses in a HW when connected - Can change RPC of default ethereum network github.com/0xngmi/metamask-e…
69
364
47
1,529
🚨 MetaMask Tracking IP Address 🚨 Consensys, the company that owns MetaMask, just updated its Privacy Policy to state that they are collecting your IP ADDRESS and your ETH ADDRESS when you send a transaction. Source: consensys.net/privacy-policy Your options 🧵 ↓ 1/4
264
1,050
358
2,465
aramanja retweeted
We’re excited to share, we have: 1) Raised $200M and now have a total of $458M in mission-driven funding 2) Decided to go FULL OPEN SOURCE at our next milestone 3) Reserved ⅔ of a potential future token for the ecosystem 4) Formed a partnership with @OpenZeppelin 👇1/8
8,420
266,502
3,124
269,313
✨🔴QiDao @optimismFND x Galxe Quest is live! 🔴✨ 1⃣ Learn about $MAI and mai.finance 2⃣Test your knowledge 3⃣Complete the quest 🏆 Get $QI / $OP airdropped and a @GalxeHQ NFT! 👉 Start here: app.optimism.io/quests/qidao…
We’re always looking for opportunities to make Optimism more accessible for both new & experienced users alike. That's why we're excited to announce Optimism Quests: a guided exploration of the ecosystem that allows users to earn NFTs while exploring Optimism apps!
1,051
21,484
291
22,745
aramanja retweeted
Baby Alpha is here! We are proud to announce that today we hit our Baby Alpha milestone and have begun the march to Fair Onboarding Alpha and Full Launch Alpha. blog.matter-labs.io/baby-alp… What we've accomplished and what's ahead 🧵 (1/6)
6,720
280,589
2,687
282,085
Make no mistake about it - Apple is now the largest threat to Web3 with their most recent App Store Guidelines that they published today. Apps Are Bullshit Part Deux. A thread below on the implications this has for Web3 & Crypto 🧵
115
371
155
1,594
GM FELLOW DEGENS nftperp is proud to announce its first-ever Paper Trading Competition! Trade, compete, and earn rewards such as Milady Maker NFTs and ERC-20s, Good Minds NFTs, future $NFTP tokens, and much more! ⬇️⬇️⬇️
13
843
8
943
aramanja retweeted
Pixel is looking for talented degens to help with his web3 agency @Cambrian_Labs Pinging all: - Content writers(ghosts) - shitposters (SMM) - marketers This will be a paid role, contractor at first. if you shine, equity and full time work available. ❤️
13
4
28
aramanja retweeted
Started my NFT journey 365 days ago. Still doesn't feel real. Giving this nft for free. If you like it, just reply something. Rt and follow not necessary, but appreciated. opensea.io/assets/ethereum/0…
49
23
3
94