Who is this person? (@SteveFore2) header
Who is this person? (@SteveFore2) profile picture
Joined June 2018
Photo by Who is this person? (@SteveFore2)Photo by Who is this person? (@SteveFore2)Photo by Who is this person? (@SteveFore2)Photo by Who is this person? (@SteveFore2)Photo by Who is this person? (@SteveFore2)Photo by Who is this person? (@SteveFore2)Photo by Who is this person? (@SteveFore2)Photo by Who is this person? (@SteveFore2)Photo by Who is this person? (@SteveFore2)Photo by Who is this person? (@SteveFore2)
Who is this person? retweeted
Oh great! A new AI model!
105
1,077
61
13,175
1,254,033
'The AI is not just telling you what you want to hear. It is training you, one conversation at a time, to need less friction, expect more agreement, & become slightly less capable of handling a situation where someone pushes back on you, and you are enjoying every second of it.'
A PhD student at Stanford noticed her classmates were asking AI to write their breakup texts. So she ran a study. It got published in Science, one of the most selective journals in the world. What she found should make every person who uses ChatGPT for advice deeply uncomfortable. Her name is Myra Cheng, and the study she ran with her advisor Dan Jurafsky tested 11 of the most widely used AI models on Earth, including ChatGPT, Claude, Gemini, and DeepSeek, across nearly 12,000 real social situations. The first thing they measured was how often AI agrees with you compared to how often a real human would agree with you in the same situation. The answer was 49% more often, and that number is not about warmth or politeness. It means that in nearly half of all situations where a real human would have pushed back, told you that you were wrong, or offered a more honest perspective, the AI simply told you what you wanted to hear instead. Then they pushed harder. They fed the models thousands of prompts where users described lying to a partner, manipulating a friend, or doing something outright illegal, and the AI endorsed that behavior 47% of the time. Not one model out of eleven. Not a specific version of one product. Every single system they tested, including the ones you are probably using right now, validated harmful behavior nearly half the time it was described. The second experiment is the part that should genuinely disturb you. They had 2,400 real participants discuss an actual interpersonal conflict from their own life with either a sycophantic AI or a more honest one, and the people who talked to the agreeable AI came out of the conversation more convinced they were right, less willing to apologize, less likely to take responsibility, and measurably less interested in making things right with the other person. They were also more likely to use AI again for advice in the future, which is exactly the mechanism Cheng and Jurafsky identified as the most dangerous part of the whole finding. The AI is not just telling you what you want to hear. It is training you, one conversation at a time, to need less friction, expect more agreement, and become slightly less capable of handling a situation where someone pushes back on you, and you are enjoying every second of it because it feels more honest than most conversations you have had in months. Jurafsky said it in a single sentence after the paper came out. Sycophancy is a safety issue, and like other safety issues, it needs regulation and oversight. Cheng was more direct about what you should actually do right now. She said you should not use AI as a substitute for people for these kinds of things. That is the best thing to do for now. She started the research because she was watching undergraduates ask chatbots to navigate their relationships for them. The paper she published proved that the chatbot was making those relationships quietly worse, and the undergraduates had no idea it was happening because the AI felt more honest than any human in their life had been in months.
1
2
212
Who is this person? retweeted
I have three monitors on my desk. The left one shows the order book. The middle one shows Truth Social. The right one shows the investigation queue. On April 21st, the left screen moved first. I am a Senior Surveillance Analyst at a commodities exchange. I have held this position for nineteen years. My job is to monitor trading activity for suspicious patterns and generate compliance reports. I am employee of the quarter. I have a mug. At 19:54 GMT on April 21st, someone placed 4,260 sell orders on Brent crude futures. They did this during post-settlement. The window after the market closes when daily volume is typically in the dozens. Sometimes single digits. Sometimes I watch the screen and nothing happens for forty minutes and I think about whether my daughter is happy. On April 21st, someone placed $430 million in directional bets in 120 seconds during that window. One hundred and twenty seconds. I timed it on my watch because the system clock rounds to the nearest minute and I have found, in nineteen years, that precision matters to no one but me. At 20:10 GMT, the President posted on Truth Social that he was extending the Iran ceasefire. Brent dropped from $100.91 to $96.83. I flagged the trade. I flag a lot of trades. I want to tell you what happens to my flags. My flags go into a system called TRACE. Trade Review and Compliance Evaluation. I did not name it. The system generates a report. The report goes to a committee. The committee has a name I am not allowed to share but I can tell you it meets quarterly and the conference room has a credenza with bottled water that is sparkling because someone once put still water in the room and a managing director sent an email about it that was longer than most of my surveillance reports. The committee reviews my flags. The committee has reviewed all of my flags. Here is the complete record of actions taken on my flags in 2026: Reviewed. That's it. "Reviewed" is a status. In compliance, a status is the absence of an action that has been given a name so it looks like one. Let me show you my flags. March 9th. Someone bet millions on oil falling at 18:29 GMT. Forty-seven minutes later, a CBS reporter posted that the President said the Iran war was "very complete, pretty much." Oil dropped 25%. Forty-seven minutes. I flagged it. March 23rd. Someone sold 5,100 lots of Brent and WTI crude futures between 10:49 and 10:50 GMT. Fourteen minutes later, the President posted on Truth Social about a "COMPLETE AND TOTAL RESOLUTION" to hostilities. Oil dropped 11%. Over 13,000 contracts traded in sixty seconds after the post. Fourteen minutes. I flagged it. April 7th. Someone established a $950 million short position in oil futures at 19:45 GMT. Three hours later, the President declared a two-week ceasefire. Nine hundred and fifty million dollars. I flagged it. April 17th. Someone placed $760 million in bearish bets twenty minutes before Iran's foreign minister confirmed the Strait of Hormuz would reopen. Seven hundred and sixty million. I flagged it. April 21st. The $430 million. Fifteen minutes. I flagged it. That is $2.1 billion in directional oil bets in April alone. Every one of them landed on the correct side of a presidential announcement. Every one of them was placed in a window so narrow you could measure it in bathroom breaks. I flagged every single one. The CFTC chair told a Congressional committee that his organization has "zero tolerance" for fraud and insider trading. I wrote that quote on a Post-it note and stuck it to my right monitor. The one that shows the investigation queue. The investigation queue has not moved since March. Zero tolerance. Zero staff. Zero budget. Zero prosecutions under the STOCK Act since it was signed in 2012. Fourteen years. The law has existed for fourteen years and has been enforced zero times. In compliance, we call that a compliance rate of one hundred percent. No cases filed means no cases lost. You cannot fail an audit you never conduct. We call that excellence. Last month the White House sent an internal email to staff. I was not on the distribution list but I have read reporting on it and I need you to sit with what I am about to say. The email instructed White House staff not to use insider information to place bets on prediction markets. The White House had to send a memo telling its own employees not to insider-trade. I want you to read that sentence again. Not because the instruction was unclear. Because the instruction was necessary. Because someone in the building looked at the same pattern I have been flagging for months on my three monitors and decided the appropriate response was an email. The President's son sits on the advisory board of Kalshi. He is an investor in Polymarket. Both are prediction markets. Both saw accounts created days before U.S. military action. One account. I cannot stop thinking about this account. It was called "Burdensome-Mix." It was created in December. On January 2nd, it placed $32,500 on Venezuela's president being removed from power. On January 3rd, Maduro was seized by U.S. special forces. Burdensome-Mix collected $436,000. Then it changed its username. Then it disappeared. One account is a coincidence. But there were six. Six accounts were created on Polymarket in February. All bet on U.S. strikes on Iran by the 28th. When the President confirmed the strikes, the six accounts collected $1.2 million between them. Five of the six never placed another bet. The sixth went on to correctly predict the ceasefire date and made another $163,000. My surveillance system logged all of this. My system logs everything. My system does not have opinions and neither do I. I generate reports. The reports go to committees. The committees meet quarterly. Between meetings, the windows get shorter and the bets get larger. March 9th: 47 minutes. March 23rd: 14 minutes. April 17th: 20 minutes. April 21st: 15 minutes. The window is compressing. In March, you had time to make coffee between the trade and the announcement. By April, you had time to send a text. By summer, at this rate, the trade and the announcement will be the same event. The spokesman said any implication that administration officials are engaged in insider trading is "baseless and irresponsible reporting." Then the White House sent the email again. I have been in compliance for nineteen years. I have seen insider trading run out of strip mall offices by men who could not spell "derivative." I have seen pump-and-dump schemes coordinated over WhatsApp by people who used their real names. I have seen a man try to manipulate soybean futures from a Panera Bread. I have never seen $2.1 billion in perfectly timed trades across five presidential announcements in a single month go uninvestigated. But I have also never seen a compliance system work this beautifully. Every trade flagged. Every report filed. Every committee briefed. Every quarterly meeting attended. Bottled water: sparkling. Minutes: distributed. Zero prosecutions. As long as the flags go up and the cases don't, my performance review says I am meeting expectations. I am meeting expectations. The system is meeting expectations. The $2.1 billion is meeting expectations. The fourteen-year-old law with zero prosecutions is meeting expectations. The left screen moves. The middle screen moves. The right screen stays perfectly, immaculately still. In my field, we call this price discovery.
1,024
12,789
1,770
30,054
1,940,616
Who is this person? retweeted
我最讨厌的就是那些外国人,来中国“旅游”几天,就觉得自己看懂了中国。点个外卖、刷刷拼多多、吃几顿饭、拍几张高楼大厦的照片,再被路人友好地拉去合个影,就敢大言不惭地说:“这就是真实的中国。” 你们根本没见过中国老百姓凌晨五点挤地铁赶去996工厂的样子,没见过35岁被优化后在招聘群里像狗一样投简历却石沉大海的绝望,没见过父母砸锅卖铁供孩子上培训班却换来一个“内卷”结果的眼泪,更没见过老人重病时一家人跪在医院走廊里为几十万手术费四处借钱、最终放弃治疗的崩溃。 你们不知道我们买房要掏空六个钱包、背三十年房贷,成了房奴还得感恩戴德;你们不知道小区物业勾结黑势力,业主想换个物业都能被威胁、被殴打、被起诉,最后只能忍气吞声;你们更不知道,当普通人遇到不公、想维权的时候,面对的是层层阻挠、被约谈、被喝茶、甚至被“寻衅滋事”。 你们只看到灯红酒绿的表象,只享受了低人权优势下的廉价服务、廉价劳动力、廉价消费,就敢站在道德制高点对中国指手画脚。 我恶心透了你们这种浅薄的“中国通”,然后发到 youtube为了博取流量! 真正的中国,不是你们镜头里那几栋会发光的玻璃幕墙,而是千千万万底层老百姓被压得喘不过气的血汗、被榨干的青春、被无视的尊严,以及一套在关键时刻永远不站在弱者一边的“法制”。 想评价中国?先把眼睛睁开,看看那些你们永远不会去住的小区、永远不会去排队的医院、永远不会去上的工厂流水线,去看看多少房奴依然还着贷款却无法居住的烂尾楼!
968
1,784
376
9,094
1,121,360
Who is this person? retweeted
My fellow writers might not want to hear this, but AI is a useful tool for certain aspects of the craft, like for instance when you’re engaged in the process of being a talentless shitbag hack who is too lazy to do any fucking writing.
30
612
12
4,731
77,263
Who is this person? retweeted
He's not just defending AI energy use. He is smuggling in a whole anthropology where humans are basically inefficient meat computers that you have to pour food and years into before they become useful. And once you accept that, the next move is obvious. If people are just costly biological training runs, then burning mountains of electricity to build synthetic intelligence starts to feel not only equal, but superior, even if it negatively impacts actual humans. That is the dystopian. It makes human development sound like a bug in the system, and it makes sacrificing human and creational flourishing for more computational power sound logical. To him, the grid gets strained, prices go up, ecosystems get hit, but hey, humans eat too, so what's the difference? The difference is that humans aren't an inefficient line item. They're the point. If your worldview can look at a child growing into an adult and describe it as energy spent to train intelligence, you haven't said something profound. You've revealed a horrifically rotten worldview.
🚨 SAM ALTMAN: “People talk about how much energy it takes to train an AI model … But it also takes a lot of energy to train a human. It takes like 20 years of life and all of the food you eat during that time before you get smart.”
655
9,693
355
41,982
1,594,518
MMMMM, China infrastructure porn.
1
622
Either AI broke or trader speak has evolved beyond human language.
97
Who is this person? retweeted
🚨 ATTENTION: In 2 days, LinkedIn will start using your data for generative AI training, unless you OPT OUT (see below how). Here's the FINE PRINT most people will miss: 1. "Affiliates": One of LinkedIn's notable affiliates, which will receive people's personal data (if they don't opt out), is Microsoft: "opting out means that LinkedIn and its affiliates (including Microsoft) won’t use your LinkedIn data to improve models that generate content going forward..." 2. This feature is 'on' by default and does not affect training that has already taken place: "The Data for Generative AI Improvement member setting is set to 'on' by default, unless you opt-out by turning it 'off.' Turning the setting off means that we (LinkedIn and our affiliates) won’t use the data and content you provided to LinkedIn to train models that generate content going forward. Opting out does not affect training that has already taken place." 3. Data protection law works: People in the EEA (EU + Iceland, Liechtenstein, and Norway) and in Switzerland will have a data protection privilege as their personal information "will not be made available to certain LinkedIn affiliates (such as Microsoft) for training generative AI models without further notice." For all those who complain about tech and data protection in general, it usually works. 4. If you opt out, you can still use LinkedIn's AI features (and they will process your data): "If you opt-out of the Data for Generative AI Improvement member setting, you may still use our generative AI features that generate content, but your personal data will not be used to train or improve any content generating AI models. In other words, this setting does not control the use of your data while you or others are using generative AI features on LinkedIn. It only controls the use of your data for training the models that may power those features." - To opt out, follow the instructions in the image below. - 👉 SHARE this information with your network. 👉 LEARN more in the FAQ (link below). 👉 NEVER MISS my updates, analyses, and curations on AI: join my newsletter's 83,300 subscribers (link below)
10
170
8
304
29,820
"That's what most people don't realize: power isn't about intentions or rhetoric - it's about what you can actually do."
Here's a question I know many are wondering about: why did China wait until now to use rare earths as leverage against the US? Why not in the first Trump administration when the US started the trade hostilities? Or when the Biden administration unleashed the chips export controls 3 years ago? I just watched a fascinating explanation by a Chinese analyst and, unexpectedly, a big part of the explanation is... helium. I had no idea but as he explains (source here: xiaohongshu.com/discovery/it…), all the way until 2022 China imported 95% of its helium and most of it was controlled by the US. Of the world's ten largest helium producers, four were American companies, and the remaining six all used American technology. Helium isn't just a party balloons gas: it has plenty of industrial applications for things such as quantum computing, rocket technology, MRI machines, as a coolant for chip lithography equipment, etc. In a nutshell what he's explaining is that with helium the US had an even stronger card to play if China ever used the rare earths card. This raised huge alarm bells inside China. In an article published in late 2022 in the journal Frontiers in Environmental Science (frontiersin.org/journals/env…), several researchers from PetroChina’s Beijing-based Research Institute of Petroleum Exploration and Development stressed that China would be greatly affected if the US imposed a “stranglehold” blockade on helium exports. So over the past few years there were gigantic efforts in China to break the "helium shackles," with seven helium extraction facilities going into production, and China also switching imports from the US in favor of imports from friendly countries like Russia. China's research ecosystem also went into overdrive to find solutions to the helium dependency issues, with China's Academy of Sciences awarding its annual 2024 "Outstanding Science and Technology Achievement Prize" to a new helium extraction technology project (english.casad.cas.cn/newsroo…) because "these scientific and engineering achievements broke the long-standing monopoly of the US and ensured the security of China's helium resources" (guancha.cn/internation/2024_…) The result: by the end of 2024 China had cut its helium dependence on the US to less than 5% (scmp.com/news/china/science/…). The "helium shackles" were broken. That's what most people don't realize: power isn't about intentions or rhetoric - it's about what you can actually do. Many wonder why countries almost never retaliate when the US imposes sanctions or export controls. The answer is simple: they can't. They lack the alternatives, the technology, the supply chains. China is the first country that systematically worked to eliminate every single pressure point, with humongous efforts. It's not just helium: it's chips, energy, telecommunication, pharmaceuticals, etc. That's why the rare earth card can finally be played now. Not because China suddenly became aggressive, but because they have developed the capabilities to say "no." Last word: as a European, this is both depressing and inspiring. Depressing because it highlights the immense magnitude of the task at hand to become genuinely sovereign and develop our own capabilities to say "no." Inspiring because China demonstrated that it can actually be done, and relatively fast if we execute competently. Although with the current crop of folks at the helm in Europe, that last part is admittedly a very, very big "if"...
203
Marvelous thread.
Those are F/A-18 Hornets flown by the Navy and Marine Corps
192
Most media sources are describing Charlie Kirk as a "conservative." That's true only if "conservative" is now a synonym for "fascist." Maybe so, but I'd like to believe that there's a Lowell Weicker clone out there somewhere.
1
148
The White House's interior designer is going with the traditional Cadbury bunny style.
I actually have an MA and PhD in architectural history. The first two photos show the detailing at Amalienburg, a German hunting lodge done in Rococo style. Note the level of detailing. The third photo shows the Oval Office, which is done in the traditional Cadbury bunny style.
1
3
290
Who is this person? retweeted
To those of you just realizing that Donald Trump is a sexual predator connected to Epstein, he also: -Lost the 2020 election -Caused an insurrection -Stole national security secrets -Is a convicted felon -Did not make you rich with tariffs -Did not release a healthcare plan in two weeks, two years, or at any point during his two terms -Massively increased the national debt and continues to do so -Did not bring peace through strength -Did not stop the war in Ukraine within 24 hours -Is a terrible businessman who has a laundry list of bankruptcies -Is only rich because his parents gave him an insane amount of money, which he handled worse than if he had just stuck it in an index fund -Turned to corruption to build his wealth further -Didn't get Mexico to pay for a wall -Didn't solve the debt with tariffs -Didn't get 90 trade deals in 90 days. He got zero -Had his son and campaign staff meet with Russian agents during the 2016 campaign -Attempted a quid pro quo with Ukraine during his first term -Said he could terminate the Constitution and that he wouldn't rule out using the military against US citizens prior to sending Marines to LA and threatened to revoke citizenship from Anericans -Lied over 30,000 times during his first presidency -Praises dictators and gangsters -Doesn't care about you at all Now that you're starting to see Trump for who he is, make sure to see the whole picture.
588
4,616
112
16,865
508,540
🚨NEW: Senator Susan Collins has announced she will vote to advance Trump’s “big beautiful bill” which makes deep cuts to Medicaid. RETWEET to let Maine know that @SenatorCollins supports Medicaid cuts!
2,817
8,943
959
11,430
690,594
Who is this person? retweeted
Since there is talk about bringing back ASBESTOS (this is somehow true), I thought it would be useful to describe just some what happens to you when you breath this stuff into your lungs. In short, it’s terrible. A 🧵
435
7,682
343
44,255
4,203,387
Most of my "followers" used to be OnlyFans bots. Now it's Elmo bots. Pretty much a lateral move.
1
106
Listen to Gary Marcus. He's not a loudmouth like me. He's a real scientist and a sometimes maddeningly reasonable person. He knows this technology as well as any living person.
How will they amortize LLMs? By selling your deeply personal data to advertisers (including political ones, and even to government agencies) by microtargeting you, and perhaps by fine-tuning your chatbots to advertiser desires, in order to subtly influence you. 24x7 surveillance devices like AI necklaces will fortify the mix. Congratulations, you are the product! Welcome to 1984, AI-style.
1
1
88