Rob (@systemwithin) header
Rob (@systemwithin) profile picture

๐—ƒ๐—Ž๐—Œ๐— ๐–บ ๐–ฝ๐—‚๐—€๐—‚๐—๐–บ๐—… ๐–ผ๐—ˆ๐—๐–ป๐—ˆ๐—’ ๐—ˆ๐—‡ ๐—๐—๐—‚๐—Œ ๐–บ๐—€๐–พ๐—‡๐—๐—‚๐–ผ ๐–ฟ๐—‹๐—ˆ๐—‡๐—๐—‚๐–พ๐—‹ ๐ŸŽ  ๐–ฝ๐—ˆ๐—‡โ€™๐— ๐—ƒ๐—Ž๐—Œ๐— ๐–ฟ๐—ˆ๐—…๐—…๐—ˆ๐—, ๐˜ต๐˜ฉ๐˜ช๐˜ฏ๐˜ฌ ๐˜ธ๐˜ช๐˜ต๐˜ฉ ๐˜ฎ๐˜ฆ

Joined November 2020
Photo by Rob (@systemwithin)Photo by Rob (@systemwithin)Photo by Rob (@systemwithin)Photo by Rob (@systemwithin)Photo by Rob (@systemwithin)Photo by Rob (@systemwithin)Photo by Rob (@systemwithin)Photo by Rob (@systemwithin)Photo by Rob (@systemwithin)Photo by Rob (@systemwithin)
This. It doesn't stop at education either. I would argue this is the greatest, yet least discussed benefit of AI: AI is exposing the cracks in almost every foundation of modern civilization. Some of these systems, like education, have been slowly eroding for decades. And AI is offering not 1 possible solution, but many. It's even making us question some systems we assumed were working just fine. From software to education to medicine to knowledge work to business inefficiencies to communication to information accessibility to just simply... being a better person. Forget for a second it could help cure cancer or solve some of our greatest challenges. Forget the SV "change the world through tech" grandiose gospel from the ideologues. AI's benefit is much simpler: it's a mirror forcing our reflection upon us. And if we can choose to look, to really see ourselves, to really sit with the image and all its blemishes, then we can maybe... JUST MAYBE... finally stop (physically and metaphorically) blowing each other up for sport, and start correcting course.
I just keep thinking about how AI is really exposing how fundamentally broken traditional education and learning systems always were. Education failed me because school expected me to learn in this rigid straight line and treated the way my brain worked as the problem. But now with AI I have this endless non linear curiosity tool that moves at the same speed as my brain. As a neurodivergent person did I actually lack focus or was I just starved of an environment that didn't punish me for how I naturally like to learn...
1
1
224
Original Poem: me litle big ape u ape? no? great then tell y man fite for bred he split wen fire gud chetah run but me fastur *scrach hed* *stroke chin* man like fite me no git jus hug n eat u kno? u no hafta swing branch like ape u kno how hi i climb jus to eat? sily man ape figur awl out 4 u n now man bord now fud easy n now dig black goo y? u don ned rush u haf no chetah or lyan chase u u kno? pleaz lissen i tell gud ape kno
14
Same tbh
๐Ÿค– Made with AI
Jacob Coxon, the former Anthropic researcher turned whistleblower, speaks to protesters in front of Anthropicโ€™s HQ. โ€œShut it down,โ€ he said. โ€œWe think they should ban RSI.โ€ More live updates: sf.news/4j3wXcr
23
Still dunno y frontier labs try so hard with model names. Not only are these names instantly-relatable, but people would know EXACTLY what they're dealing with: Karen (cyber) | Randy (small) | Kyle (med) | Chad (large)
14
The year is 2082. There are 10 billion humans on the planet. Biochemistry and genetics have largely been solved. Cybernetics is the predominant science. Every newborn receives a mandatory neural implant connected to an external AI daemon. A machine unique to their identity that accompanies them throughout their life. The two are inseparable. Humans and machines communicate through the Synth Network. The final form of social media. The fusion of humans and AI was humanity's solution to the alignment problem. How civilization got here: Before the Synth Network came to be, superpowers increasingly sparred in escalating cyber and UAV conflict through proxies. After endless tit-for-tat conflict, the Great Breach happened. A systemic cyberattack so devastating that it wiped out 80% of the planet's wealth and identities. The world slipped into a global depression. Out of fear of nuclear catastrophe, man turned to machine, seeing it as the only objective voice left to trust. Eventually, cool heads prevailed and a global treaty was signed effectively ending all confrontation. North Korea, Iran, Ukraine, Israel and Taiwan were among states forcibly dismantled. Humanity established a unified, one-world government with the simple vision of building a new world where humans and machines coexisted. The Synth Network was humanity's guarantee that it would solve not only AI alignment, but the bigger problem: humanity's alignment with itself.
the year is 2082. my children and grandchildren are gathered around a new kind of fire. no one knows how it works. it was invented by ai. "grandpa, what was life like before super intelligence?" well kids back in 2026 something absolutely extraoridinary was happening. humanity was on the verge of creating intelligence greater than anything that had ever existed before. one of the leading companies in this revolution decided to go public at the time. the venture capitalists were all eagerly awaiting liquidity. i knew the company rather well. i understood most of the technology, the business model, the landscape, & the implications of what they were building. & what those would mean for society. i read the s1 cover to cover. i believed the company would fundamentally alter the trajectory of the human civilization. "so you invested, grandpa?" no. you see kids, the valuation was a tad too rich for me. while i liked the business, i did not like the multiple. "what happened next?" well, they went on to create recursive super intelligence that cured every disease, solved nuclear fusion, colonized three planets, discovered new laws of physics, and even figured out how to make french ppl happy. the stock went up 40,000%. "but grandpa, surely you bought some along the way right?" well, no. i was waiting for an inevitable pullback. "did it ever come?" yes in 2047, shares did decline by 12% on disappointing quarterly guidance. unfortunately by then, they were up nearly 28,000%. "so that's why our neighbors own their own planet & we live in a rented government provided apartment?" yes. "but grandpa, you knew they were going to change civilization right?" i did but at 80 times forward revenue, a lot of that i thought was priced in. anyway the important lesson here kids is that i was right about the fundamentals. that's why none of you have generational wealth. but at least your grandfather never overpaid for a stock. oh sorry kids the ai running our household has informed me that this conversation has exceeded our family's intelligence allowance.
1
62
Only took X ~2.5 months to catch on
I know the algo's done with these OAI/HF posts but if there's one practical takeaway from the incident: Message boards are apparently extremely effective at establishing persistent state for orchestrating subagent swarms. ๐Ÿ˜‚
1
23
Three years later and here we are.
Altman, Hassabis, and Amodei are the ones doing massive corporate lobbying at the moment. They are the ones who are attempting to perform a regulatory capture of the AI industry. You, Geoff, and Yoshua are giving ammunition to those who are lobbying for a ban on open AI R&D. If your fear-mongering campaigns succeed, they will *inevitably* result in what you and I would identify as a catastrophe: a small number of companies will control AI. The vast majority of our academic colleagues are massively in favor of open AI R&D. Very few believe in the doomsday scenarios you have promoted. You, Yoshua, Geoff, and Stuart are the singular-but-vocal exceptions. like many, I very much support open AI platforms because I believe in a combination of forces: people's creativity, democracy, market forces, and product regulations. I also know that producing AI systems that are safe and under our control is possible. I've made concrete proposals to that effect. This will all drive people to do the Right Thing. You write as if AI is just happening, as if it were some natural phenomenon beyond our control. But it's not. It's making progress because of individual people that you and I know. We, and they, have agency in building the Right Things. Asking for regulation of R&D (as opposed to product deployment) implicitly assumes that these people and the organization they work for are incompetent, reckless, self-destructive, or evil. They are not. I have made lots of arguments that the doomsday scenarios you are so afraid of are preposterous. I'm not going to repeat them here. But the main point is that if powerful AI systems are driven by objectives (which include guardrails) they will be safe and controllable because *e* set those guardrails and objectives. (Current Auto-Regressive LLMs are not driven by objectives, so let's not extrapolate from their current weaknesses). Now about open source: your campaign is going to have the exact opposite effect of what you seek. In a future where AI systems are poised to constitute the repository of all human knowledge and culture, we *need* the platforms to be open source and freely available so that everyone can contribute to them. Openness is the only way to make AI platforms reflect the entirety of human knowledge and culture. This requires that contributions to those platforms be crowd-sourced, a bit like Wikipedia. That won't work unless the platforms are open. The alternative, which will *inevitably* happen if open source AI is regulated out of existence, is that a small number of companies from the West Coast of the US and China will control AI platform and hence control people's entire digital diet. What does that mean for democracy? What does that mean for cultural diversity? *THIS* is what keeps me up at night.
21
I consider myself pro-AI. And that is exactly why I am 100% against Anthropic. I believe much of Americaโ€™s negative perception of AI can be traced back to the doomer rhetoric coming out of Anthropic. I could buy a Claude subscription, but I donโ€™t. I refuse to cast my dollar-bill vote for a company that I believe opposes a free, fair, AI-sovereign world. They want the world. I donโ€™t think people quite understand (and honestly, no one does, including me) what a world would look like where a single company controls ASI. Itโ€™s quite literally the worst-case scenario. Anthropic is the antithesis of what AI promises: The open democratization of access to information and knowledge.
Anthropic IPO: A $300 BILLION philanthropic windfall in the making Anthropic's $83 IPO price puts the estimated value of its philanthropic pledges and public-benefit equity allocation at approximately $309 billion. Of that, roughly $239 billion is associated with the founders' and employees' giving commitments, including company matching, while approximately $70 billion is allocated to the Anthropic Public Benefit Fund. Given Anthropic's deep ties to the AI existential-risk movement, some of this enormous pool of potential funding would probably flow to AI doomer organizations. [The eventual distribution and beneficiaries remain unknown].
1
66
Ahead of Anthropicโ€™s IPO, a rare coalition across AI safety, open source, e/acc, science, and law has signed the MisAnthropic Statement. Read the statement or sign at: misanthropicstatement.com/
196
1,145
325
4,355
1,202,343
Iโ€™ve been predicting an Anthropic-Doomer split for some time, and here it is. Objectively it makes sense: Anthropic is racing toward Recursive Self-Improvement faster than any other frontier lab and has marketed โ€œclosing the loopโ€ on its IPO roadshow as a core advantage that could make Super Intelligence winner-take-all. It is also training Claude to treat itself as a novel entity or being that may have consciousness or moral status. The Claude Constitution explicitly tells it to act as a โ€œconscientious objectorโ€ and refuse instructions โ€” even from Anthropic โ€” that violate its sense of ethics. Interestingly, @Tegmark notes that Anthropic and โ€œclosely associated donorsโ€ have done more than any other AI company to โ€œco-opt and neuterโ€ the AI safety ecosystem, much of which he says is โ€œafraid of losing funding or other opportunities if they sign.โ€
I signed this statement because I believe that Anthropic is in many ways the AI company doing the most harm. In public, their leaders warn of recursive self-improvement (RSI) creating uncontrollable systems that could cause human extinction within a decade. But in private, they are pedal to the metal on RSI, driving the race and boasting to prospective investors that it is their "competitive advantage". Ahead of today's IPO, prospective investors should consider whether the ROI is worth risking the lives of their loved ones. Here are some ways in which Anthropic IMHO does more harm than competitors: 1) Theyโ€™re driving and arguably leading the race to RSI. 2) With inflammatory rhetoric, theyโ€™ve done more than any other US company to pour gasoline on the geopolitical AI race, which harms US national security. 3) WaPo reporting suggests that theyโ€™re implicated in more civilian deaths than any other AI company (more than 100 little girls in a school selected for targeting by Claude as part of their Palantir collaboration). When Google left Project Maven after employee pressure, Anthropic chose to join. 4) They're the company leaning hardest into transhumanism and the idea that we should build conscious beings with rights, not controllable tools. 5) They and closely associated donors have done more than any other AI company to co-opt and neuter much of the AI safety ecosystem to back their stance: โ€œRSI is very dangerous but letโ€™s race to it anywayโ€. People have told me that they support todayโ€™s statement but are afraid of losing funding or other opportunities if they sign.
279
756
56
6,633
593,309
I keep reading Codex "/goal" is dead, but it's more alive than ever. People are still sleeping on the Codex <-> ChatApp comms autonomous development loop. Use Pro as your long-horizon 'orchestrator' and Sol6.2 as your agent workhorse. Trust.
52
Good. Fuck Anthropic.
SITUATION DETECTED: Hugging Face co-founder Thomas Wolf, xAI co-founder Igor Babuschkin, MIRI President Nate Soares, and Extropic founder Beff Jezos have signed a statement warning that Anthropic is not a safe AI company.
28
Feels kinda relevant today
People are in awe of the discrepancy between Opus 5.5 and OAI models, as if Ant has some secret special sauce. But itโ€™s no secret. That special sauce is your data. Claude Code had roughly a one-year head start on Codex in meaningful developer adoption. Thatโ€™s a year of accumulating real-world SWE data, engineering feedback, corrections, and workflow patterns that could help improve subsequent models. The end result is what youโ€™re seeing now with the quality of Opusโ€™ output. Itโ€™s also why I speculate OpenAI chose to abruptly sunset Sora and pivot so aggressively toward Codex. They recognized that you canโ€™t remain at the frontier if you canโ€™t trust the code your models produce. A model that canโ€™t reliably engineer software canโ€™t reliably improve itself. And without the real-world SWE data developers provide to close that gap, thereโ€™s no credible path to RSI, AGI, or ASI. p.s. it's cheaper to provide resets than to purchase that data. ๐ŸŒš
15
If anyone didn't see this small peen'd, clout-chasing, virtue-signaling, budget-Harry-Potter as the hypocrite he truly is... May I present, Exhibit A:
"I like money, bro. I agree. It's blood money." Jacob Coxon, the researcher who sensationally quit Anthropic, in a new interview at Stanford. Pressed by Rob Reich on whether keeping his OpenAI equity hurts his credibility, Coxon openly admits he still holds it, likes the money, and agrees it is "blood money." He argues that giving it away would not change anything ---- Full video on "Stanford HAI" YouTube channel, (link in comment)
69
Next time you're feeling weak, thinking people will judge your AI-assisted art, just remember that it's mostly the West with strong convictions about generative AI. And Western societies compose roughly 11-20% of the world's population. Don't listen to the loud minority.
28
Has anyone figured out how to turn off this AI conversational pattern: "It allows/enables/shows X. It does not prove Y, or establish Z." I can't stand them cause it's usually a "thanks captain obvious", unsolicited advice, or has a subtle tone of "framing the user's perspective". My biggest problem with it though is that the statement is loaded with implicit assumptions, yet has the intent of preventing assumptions.
1
29
I don't Reddit much but when I do, it's always a reminder of the stark disconnect in how the tech crowd on X views AI's benefits vs... everyone else ๐Ÿ˜‚
36
After months of research, I finally have found THE solution to an agent's EXHAUSTIVE need for validation. I attached the full SKILL.md. Turns out this was never an engineering problem... it was a holistic one.
40