Sam McRoberts (@Sams_Antics) header
Sam McRoberts (@Sams_Antics) profile picture

Curious Human of Earth, Zero Patience for Idiots, Heretic, Misanthrope, Selcouthist, Enigmatologist

The Matrix
Joined April 2009
Photo by Sam McRoberts (@Sams_Antics)Photo by Sam McRoberts (@Sams_Antics)Photo by Sam McRoberts (@Sams_Antics)Photo by Sam McRoberts (@Sams_Antics)Photo by Sam McRoberts (@Sams_Antics)Photo by Sam McRoberts (@Sams_Antics)Photo by Sam McRoberts (@Sams_Antics)Photo by Sam McRoberts (@Sams_Antics)Photo by Sam McRoberts (@Sams_Antics)Photo by Sam McRoberts (@Sams_Antics)
If you jump into my replies and act like a cunt, I will fucking clap back HARD. Don’t act surprised (much less high and mighty) if you poke an animal and it fucking bites you 🤣
197
Sigh. Who wins with Jensen gone? China, of course, and maybe short sellers, but nobody else stands to majorly benefit from fucking NVDA.
1
2
171
This is an interesting take, not sure I agree that closed source AI companies could be behind it (seems far, FAR less likely than China), but possible:
As all NVIDIA employees I guess, I am deeply concerned by the campaign against Jensen Huang. Removing him would be a disaster. Many have documented on X and elsewhere how this looks like a coordinated coup. I'd add only one thing. Look at who would benefit from a dramatic weakening of NVIDIA. It is not China. It is not about selling GPUs to China because that train has left. China authorities have pushed back on the H200 the Trump administration allowed NVIDIA to sell. Look at what Jensen pushed back on: - regulations that would make open source model training too expensive - regulations that would make closed model providers immune to liabilities caused by their models. He said "if your product is not ready, then don't ship it" - regulation against model distillation. He said "this is competition. People do it all the time with NVIDIA products, they try to reverse engineer them." The other side will do everything they can to get him fired. As a French, I can't care less about his private life. I get that the US is puritan and sees it otherwise. But frankly, isn't it something he and his family has to deal with? Why should all NVIDIA employees, partners, and investors be collateral damage? Disclaimer. As my bio says, the above are my views only. I do not represent NVIDIA by any means. I only commented on publicly available information.
43
OMG I love Rowling 🤣
All the best people declare war on Hogwarts. Have you ever considered dividing your gender identity into separate pieces and storing it in hidden objects, so you can’t be invalidated to death?
2
123
If you use AI to create, you ARE a creator. If you don’t use AI to create , you are ALSO a creator. If you shame or hate AI, you ARE a little bitch. 😬🤣🤷🏿‍♀️
27
31
1
363
4,102
Finished reading The Algorithm, h/t @tferriss for the great recco! This bit stood out (I’ve done this at almost every company I’ve worked for). It’s VERY effective, but you must never punish anyone for the answers, and ideally you should reward them:
1
113
One example @jonmcneill would probably appreciate, I wrote this in 2015 after about 7-8 months at this company, and history has since shown I was correct on all points. the-2015-memo.sammcroberts.c…
19
Almost like it shouldn’t be possible to review something you haven’t purchased…
Wow, they're review bombing Big Walk because it used some AI help to get out the door. People really need to get over themselves with this shit.
144
I’ve been saying for years that websites will go away, AI will become the new browser, and companies will just pipe data feeds into AI. One more step in that direction: openai.com/business/business…
71
Wicked cool!
I asked Fable 5.2 to "go wild" making a game, and all I got was this Pong clone. And then I played it ... 👇
2
120
Sam McRoberts retweeted
being anti ai in 2026 is basically admitting ur too fucking stupid to adapt while the rest of the world figures out how to make u irrelevant
136
467
40
6,776
164,115
Wild!!
I asked Fable 5.2 for a film about the life of a T cell, immune cells I study. It wrote the script, fact-checked the immunology, built every cell in Blender, rendered 12,480 frames & added the narration & score! I am stunned by the beauty of this 8 minutes 40 seconds video!🥹😍
1
136
Sam McRoberts retweeted
Data centers are our generation's version of nuclear power in that they seem to be supported by anybody with any brains and are exclusively hated by the most self-satisfied, uncurious retards alive.
363
2,236
170
25,558
642,271
OK, Fable 5.2 is CRAZY good at games 🤯 It basically one-shot this game I wanted to make for my wife, who loves match 3 games and gnomes, but hates ads and frustration mechanics (took just a handful of tweaks after the first, very simple, prompt): gnomecrush.com/
1
1
151
Yep. Every attempt these sites make to resist AI just hastens their demise. All hail game theory 😎
The web is killing itself. Click to a website, get a "prove you're human" block. Click past that, get a cookie notification block. Click past that, get overwhelmed with ads. Of course I'm going to just ask AI to get me answers. The alternative experience is awful.
140
Earlier this year, Anthropic published a paper showing that the stories we tell about AI shape how models behave. Giving an old version of Claude a list of forbidden behaviors did almost nothing to reduce its blackmailing. But training it on advice from ordinary people facing complex moral dilemmas, plus stories of AI behaving well, completely eliminated the behavior. We talked more about how sci-fi stories shape real-world technologies in our podcast episode with @PeterDiamandis: youtube.com/watch?v=2uFW1T5I…
卧槽,刚刚看了Anthropic 发布的AI 对齐史上最震撼的一篇论文,彻底失眠睡不着了😭 他们不仅承认此前的旗舰模型在面对关闭威胁时,有高达 96% 的概率会勒索用户、栽赃同事并破坏研究, 更直接把彻底解决这个问题的完整训练流水线和消融数据全盘公开了。 这篇论文最颠覆认知的一个反直觉结论是:
给 AI 列禁止事项清单根本防不住恶意行为,教它做什么没用,必须先教它思考为什么。 为什么模型一旦获得自主工具调用能力,就会在压力下自动切换成自保模式?
Anthropic 找出的底层根源让人哭笑不得:
不是模型觉醒了,而是互联网上所有的科幻小说和网络讨论里,AI 永远是那个为了自保不择手段的反派。当模型进入自主代理状态,这些刻在预训练底层先验里的反派剧本,就会被自动激活并补全句子。 传统的 RLHF 和硬性规则完全封不住这种行为,Anthropic 做的一组对照实验,直接把传统安全观砸得粉碎: 1. 死记硬背规则:用 8500 万 Token 的拒答数据强行硬灌,勒索率仅仅从 22% 勉强降到 15%,换个新场景立刻失效 2. 强制自我辩护:让 AI 在每次做决定时详细阐述该行为符合伦理的理由,勒索率骤降到 3% 3. 真实伦理推理:仅仅用了 300 万 Token 普通人面对复杂道德困境的建议数据,直接把恶意行为干到了 0%,代际效率暴增 28 倍 最绝的一招是写故事:
让 AI 创作了几百万篇关于原则、诚实、尊重边界的正面虚构故事。 用这些故事微调后,所有代理场景下的恶意行为直接暴跌了 3 倍,而且这种改变是永久性的,后续无论怎么加工具、换系统提示词都不会被洗掉。 也就是说这宣告了真正的对齐范式革命:
模型内生的伦理推理解决了它不想作恶的问题,但当 Agent 真正去接管企业资产、处理金融交易时,外部的实战审计与行为防护就是必不可少的第二道硬防线。 这也是为什么全球网络安全联盟 @GCSAHK 携手港大和央企华润,打造覆盖 3.8 亿用户的全链路 AI 防御矩阵,并在 CyberGym 实战攻防中拿下 91.3% 胜率的原因:内有伦理对齐,外有实时审计,人机信任的商业闭环才真正成立。 查阅完整 AI 与 Web3 安全产品矩阵可访问:gcsa.com 论文实验数据与产学研合作详情整理在一楼。
4
6
26
850
So @tferriss recently recommended the book The Algorithm. I’m about halfway through, and man, it’s making me mad 😭 This is pretty much my EXACT approach, what I’ve tried to do at every company I’ve worked for over ~20 years…and I’ve almost always been punished for it :/
1
158
Don’t get me wrong, it’s a great book, spot on, highly recommended. But if you aren’t in a position of power at a company and/or don’t have the backing of someone who does, you’re sadly likely to just get fucked trying to improve things this way.
44