9u50fv – Page 14

Anti-vaccine group founded by RFK Jr. sues RFK Jr. over vaccine task force

CDC, children's health defense, health, HHS, Robert F. Kennedy, vaccines / 9u50fv / July 24, 2025

Now that Kennedy has moved on to politics, stepping down from his role at Children’s Health Defense (CHD) and joining the Trump administration, CHD has not let go of the issue.

Ray Flores, senior outside counsel to CHD, filed the lawsuit, which is being funded by CHD. In it, Flores notes that on March 15, 2025, he sent Kennedy a 60-day notice about the task force issue, and Kennedy did not respond.

Overall, the lawsuit contains anti-vaccine talking points and false claims, such as that childhood vaccines have not gone through safety testing (they have). Flores justifies the lawsuit saying that, without the task force, he “and his family can’t make informed decisions in light of the onslaught of current and seemingly never-ending outbreaks.”

In a social media post from CHD on Tuesday, Flores criticized Kennedy directly. “Why is he not dealing with vaccines? This is not the Bobby we know,” he said in the posted video. “Is he being held captive in the swamp? And it kind of feels that way sometimes, doesn’t it?”

It remains unclear why Kennedy has not set up the task force. HHS did not immediately respond to a request for comment from Ars Technica.

Otherwise, Kennedy has not shied from unilaterally rolling back access to vaccines and continuing to spread anti-vaccine misinformation as the country’s top health official. His hand-selected vaccine advisory committee has already announced its intention to question the entire childhood vaccine schedule.

Potential explanations

However, there is one clear detail that could potentially explain Kennedy’s delay. The 1986 law that sets up the task force is specific about who should be on it. The task force “shall consist of the Director of the National Institutes of Health, the Commissioner of the Food and Drug Administration, and the Director of the Centers for Disease Control [and Prevention],” the law reads. Currently, the CDC has no director.

Anti-vaccine group founded by RFK Jr. sues RFK Jr. over vaccine task force Read More »

White House unveils sweeping plan to “win” global AI race through deregulation

AI, AI and work, AI development tools, AI ethics, AI infrastructure, AI law, AI regulation, AI research, AI security, biosecurity, Biz & IT, china, Chips Act, data centers, David Sacks, Deepfakes, Department of Commerce, Donald Trump, Energy, Export controls, machine learning, Michael Kratsios, national security, NIST, Open Source, Policy, semiconductors, White House / 9u50fv / July 24, 2025

Trump’s plan was not welcomed by everyone. J.B. Branch, Big Tech accountability advocate for Public Citizen, in a statement provided to Ars, criticized Trump as giving “sweetheart deals” to tech companies that would cause “electricity bills to rise to subsidize discounted power for massive AI data centers.”

Infrastructure demands and energy requirements

Trump’s new AI plan tackles infrastructure head-on, stating that “AI is the first digital service in modern life that challenges America to build vastly greater energy generation than we have today.” To meet this demand, it proposes streamlining environmental permitting for data centers through new National Environmental Policy Act (NEPA) exemptions, making federal lands available for construction and modernizing the power grid—all while explicitly rejecting “radical climate dogma and bureaucratic red tape.”

The document embraces what it calls a “Build, Baby, Build!” approach—echoing a Trump campaign slogan—and promises to restore semiconductor manufacturing through the CHIPS Program Office, though stripped of “extraneous policy requirements.”

On the technology front, the plan directs Commerce to revise NIST’s AI Risk Management Framework to “eliminate references to misinformation, Diversity, Equity, and Inclusion, and climate change.” Federal procurement would favor AI developers whose systems are “objective and free from top-down ideological bias.” The document strongly backs open source AI models and calls for exporting American AI technology to allies while blocking administration-labeled adversaries like China.

Security proposals include high-security military data centers and warnings that advanced AI systems “may pose novel national security risks” in cyberattacks and weapons development.

Critics respond with “People’s AI Action Plan”

Before the White House unveiled its plan, more than 90 organizations launched a competing “People’s AI Action Plan” on Tuesday, characterizing the Trump administration’s approach as “a massive handout to the tech industry” that prioritizes corporate interests over public welfare. The coalition includes labor unions, environmental justice groups, and consumer protection nonprofits.

White House unveils sweeping plan to “win” global AI race through deregulation Read More »

SpaceX launches a pair of NASA satellites to probe the origins of space weather

falcon 9, launch, NASA, Science, Space, space weather, spacex, TRACERS, Vandenberg space force base / 9u50fv / July 24, 2025

“This is going to really help us understand how to predict space weather in the magnetosphere.”

This artist’s illustration of the Earth’s magnetosphere shows the solar wind (left) streaming from the Sun, and then most of it being blocked by Earth’s magnetic field. The magnetic field lines seen here fold in toward Earth’s surface at the poles, creating polar cusps. Credit: NASA/Goddard Space Flight Center

Two NASA satellites rocketed into orbit from California aboard a SpaceX Falcon 9 rocket Wednesday, commencing a $170 million mission to study a phenomenon of space physics that has eluded researchers since the dawn of the Space Age.

The twin spacecraft are part of the NASA-funded TRACERS mission, which will spend at least a year measuring plasma conditions in narrow regions of Earth’s magnetic field known as polar cusps. As the name suggests, these regions are located over the poles. They play an important but poorly understood role in creating colorful auroras as plasma streaming out from the Sun interacts with the magnetic field surrounding Earth.

The same process drives geomagnetic storms capable of disrupting GPS navigation, radio communications, electrical grids, and satellite operations. These outbursts are usually triggered by solar flares or coronal mass ejections that send blobs of plasma out into the Solar System. If one of these flows happens to be aimed at Earth, we are treated with auroras but vulnerable to the storm’s harmful effects.

For example, an extreme geomagnetic storm last year degraded GPS navigation signals, resulting in more than $500 million in economic losses in the agriculture sector as farms temporarily suspended spring planting. In 2022, a period of elevated solar activity contributed to the loss of 40 SpaceX Starlink satellites.

“Understanding our Sun and the space weather it produces is more important to us here on Earth, I think, than most realize,” said Joe Westlake, director of NASA’s heliophysics division.

NASA’s two TRACERS satellites launched Wednesday aboard a SpaceX Falcon 9 rocket from Vandenberg Space Force Base, California. Credit: SpaceX

The launch of TRACERS was delayed 24 hours after a regional power outage disrupted air traffic control over the Pacific Ocean near the Falcon 9 launch site on California’s Central Coast, according to the Federal Aviation Administration. SpaceX called off the countdown Tuesday less than a minute before liftoff, then rescheduled the flight for Wednesday.

TRACERS, short for Tandem Reconnection and Cusp Electrodynamics Reconnaissance Satellites, will study a process known as magnetic reconnection. As particles in the solar wind head out into the Solar System at up to 1 million mph, they bring along pieces of the Sun’s magnetic field. When the solar wind reaches our neighborhood, it begins interacting with Earth’s magnetic field.

The high-energy collision breaks and reconnects magnetic field lines, flinging solar wind particles across Earth’s magnetosphere at speeds that can approach the speed of light. Earth’s field draws some of these particles into the polar cusps, down toward the upper atmosphere. This is what creates dazzling auroral light shows and potentially damaging geomagnetic storms.

Over our heads

But scientists still aren’t sure how it all works, despite the fact that it’s happening right over our heads, within the reach of countless satellites in low-Earth orbit. But a single spacecraft won’t do the job. Scientists need at least two spacecraft, each positioned in bespoke polar orbits and specially instrumented to measure magnetic fields, electric fields, electrons, and ions.

That’s because magnetic reconnection is a dynamic process, and a single satellite would provide just a snapshot of conditions over the polar cusps every 90 minutes. By the time the satellite comes back around on another orbit, conditions will have changed, but scientists wouldn’t know how or why, according to David Miles, principal investigator for the TRACERS mission at the University of Iowa.

“You can’t tell, is that because the system itself is changing?” Miles said. “Is that because this magnetic reconnection, the coupling process, is moving around? Is it turning on and off, and if it’s turning on and off, how quickly can it do it? Those are fundamental things that we need to understand… how the solar wind arriving at the Earth does or doesn’t transfer energy to the Earth system, which has this downstream effect of space weather.”

This is why the tandem part of the TRACERS name is important. The novel part of this mission is it features two identical spacecraft, each about the size of a washing machine flying at an altitude of 367 miles (590 kilometers). Over the course of the next few weeks, the TRACERS satellites will drift into a formation with one trailing the other by about two minutes as they zip around the world at nearly five miles per second. This positioning will allow the satellites to sample the polar cusps one right after the other, instead of forcing scientists to wait another 90 minutes for a data refresh.

With TRACERS, scientists hope to pick apart smaller, fast-moving changes with each satellite pass. Within a year, TRACERS should collect 3,000 measurements of magnetic reconnections, a sample size large enough to start identifying why some space weather events evolve differently than others.

“Not only will it get a global picture of reconnection in the magnetosphere, but it’s also going to be able to statistically study how reconnection depends on the state of the solar wind,” said John Dorelli, TRACERS mission scientist at NASA’s Goddard Space Flight Center. “This is going to really help us understand how to predict space weather in the magnetosphere.”

One of the two TRACERS satellites undergoes launch preparations at Millennium Space Systems, the spacecraft’s manufacturer. Credit: Millennium Space Systems

“If we can understand these various different situations, whether it happens suddenly if you have one particular kind of event, or it happens in lots of different places, then we have a better way to model that and say, ‘Ah, here’s the likelihood of seeing a certain kind of effect that would affect humans,'” said Craig Kletzing, the principal investigator who led the TRACERS science team until his death in 2023.

There is broader knowledge to be gained with a mission like TRACERS. Magnetic reconnection is ubiquitous throughout the Universe, and the same physical processes produce solar flares and coronal mass ejections from the Sun.

Hitchhiking to orbit

Several other satellites shared the ride to space with TRACERS on Wednesday.

These secondary payloads included a NASA-sponsored mission named PExT, a small technology demonstration satellite carrying an experimental communications package capable of connecting with three different networks: NASA’s government-owned Tracking and Data Relay Satellites (TDRS) and commercial satellite networks owned by SES and Viasat.

What’s unique about the Polylingual Experimental Terminal, or PExT, is its ability to roam across multiple satellite relay networks. The International Space Station and other satellites in low-Earth orbit currently connect to controllers on the ground through NASA’s TDRS satellites. But NASA will retire its TDRS satellites in the 2030s and begin purchasing data relay services using commercial satellite networks.

The space agency expects to have multiple data relay providers, so radios on future NASA satellites must be flexible enough to switch between networks mid-mission. PExT is a pathfinder for these future missions.

Another NASA-funded tech demo named Athena EPIC was also aboard the Falcon 9 rocket. Led by NASA’s Langley Research Center, this mission uses a scalable satellite platform developed by a company named NovaWurks, using building blocks to piece together everything a spacecraft needs to operate in space.

Athena EPIC hosts a single science instrument to measure how much energy Earth radiates into space, an important data point for climate research. But the mission’s real goal is to showcase how an adaptable satellite design, such as this one using NovaWurks’ building block approach, might be useful for future NASA missions.

A handful of other payloads rounded out the payload list for Wednesday’s launch. They included REAL, a NASA-funded CubeSat project to investigate the Van Allen radiation belts and space weather, and LIDE, an experimental 5G communications satellite backed by the European Space Agency. Five commercial spacecraft from the Australian company Skykraft also launched to join a constellation of small satellites to provide tracking and voice communications between air traffic controllers and aircraft over remote parts of the world.

Stephen Clark is a space reporter at Ars Technica, covering private space companies and the world’s space agencies. Stephen writes about the nexus of technology, science, policy, and business on and off the planet.

SpaceX launches a pair of NASA satellites to probe the origins of space weather Read More »

OpenAI and partners are building a massive AI data center in Texas

AI, AI chips, AI GPU, AI infrastructure, Biz & IT, datacenters, Larry Ellison, machine learning, microsoft, NVIDIA, openai, oracle, sam altman, SoftBank, Stargate / 9u50fv / July 24, 2025

Stargate moves forward despite early skepticism

When OpenAI announced Stargate in January, critics questioned whether the company could deliver on its ambitious $500 billion funding promise. Trump ally and frequent Altman foe Elon Musk wrote on X that “They don’t actually have the money,” claiming that “SoftBank has well under $10B secured.”

Tech writer and frequent OpenAI critic Ed Zitron raised concerns about OpenAI’s financial position, noting the company’s $5 billion in losses in 2024. “This company loses $5bn+ a year! So what, they raise $19bn for Stargate, then what, another $10bn just to be able to survive?” Zitron wrote on Bluesky at the time.

Six months later, OpenAI’s Abilene data center has moved from construction to partial operation. Oracle began delivering Nvidia GB200 racks to the facility last month, and OpenAI reports it has started running early training and inference workloads to support what it calls “next-generation frontier research.”

Despite the White House announcement with President Trump in January, the Stargate concept dates back to March 2024, when Microsoft and OpenAI partnered on a $100 billion supercomputer as part of a five-phase plan. Over time, the plan evolved into its current form as a partnership with Oracle, SoftBank, and CoreWeave.

“Stargate is an ambitious undertaking designed to meet the historic opportunity in front of us,” writes OpenAI in the press release announcing the latest deal. “That opportunity is now coming to life through strong support from partners, governments, and investors worldwide—including important leadership from the White House, which has recognized the critical role AI infrastructure will play in driving innovation, economic growth, and national competitiveness.”

OpenAI and partners are building a massive AI data center in Texas Read More »

AI video is invading YouTube Shorts and Google Photos starting today

AI, Artificial Intelligence, Google, google photos, Tech, youtube / 9u50fv / July 24, 2025

Google is following through on recent promises to add more generative AI features to its photo and video products. Over on YouTube, Google is rolling out the first wave of generative AI video for YouTube Shorts, but even if you’re not a YouTuber, you’ll be exposed to more AI videos soon. Google Photos, which is integrated with virtually every Android phone on the market, is also getting AI video-generation capabilities. In both cases, the features are currently based on the older Veo 2 model, not the more capable Veo 3 that has been meming across the Internet since it was announced at I/O in May.

YouTube CEO Neal Mohan confirmed earlier this summer that the company planned to add generative AI to the creator tools for YouTube Shorts. There were already tools to generate backgrounds for videos, but the next phase will involve creating new video elements from a text prompt.

Starting today, creators will be able to use a photo as the basis for a new generative AI video. YouTube also promises a collection of easily applied generative effects, which will be accessible from the Shorts camera. There’s also a new AI playground hub that the company says will be home to all its AI tools, along with examples and suggested prompts to help people pump out AI content.

The Veo 2-based videos aren’t as realistic as Veo 3 clips, but an upgrade is planned.

So far, all the YouTube AI video features are running on the Veo 2 model. The plan is still to move to Veo 3 later this summer. The AI features in YouTube Shorts are currently limited to the United States, Canada, Australia, and New Zealand, but they will expand to more countries later.

AI video is invading YouTube Shorts and Google Photos starting today Read More »

GPT Agent Is Standing By

Agent / 9u50fv / July 24, 2025

OpenAI now offers 400 shots of ‘agent mode’ per month to Pro subscribers.

This incorporates and builds upon OpenAI’s Operator. Does that give us much progress? Can it do the thing on a level that makes it useful?

So far, it does seem like a substantial upgrade, but we still don’t see much to do with it.

Greg Brockman (OpenAI): When we founded OpenAI (10 years ago!!), one of our goals was to create an agent that could use a computer the same way as a human — with keyboard, mouse, and screen pixels.

ChatGPT Agent is a big step towards that vision, and bringing its benefits to the world thoughtfully.

ChatGPT Agent: our first AI with access to a text browser, a visual browser, and a terminal.

Rolling out in ChatGPT Pro, Plus, and Team [July 17].

OpenAI: t the core of this new capability is a unified agentic system. It brings together three strengths of earlier breakthroughs: Operator’s⁠ ability to interact with websites, deep research’s⁠ skill in synthesizing information, and ChatGPT’s intelligence and conversational fluency.

The main claimed innovation is unifying Deep Research, Operator and ‘ChatGPT’ which might refer to o3 or to GPT-4o or both, plus they claim to have added unspecified additional tools. One key tool is it claims to be able to use connectors for apps like Gmail and GitHub.

As always with agents, one first asks, what do they think you will do with it?

What’s the pitch?

OpenAI: ChatGPT can now do work for you using its own computer, handling complex tasks from start to finish.

You can now ask ChatGPT to handle requests like “look at my calendar and brief me on upcoming client meetings based on recent news,” “plan and buy ingredients to make Japanese breakfast for four,” and “analyze three competitors and create a slide deck.” ChatGPT will intelligently navigate websites, filter results, prompt you to log in securely when needed, run code, conduct analysis, and even deliver editable slideshows and spreadsheets that summarize its findings.

Okay, but what do you actually do with that? What are the things the agent does better than alternatives, and which the agent does well enough to be worth doing?

Tejal Patwardhan (OpenAI): these results were eye-opening for me… chatgpt agent performed better than i expected on some pretty realistic investment banking tasks.

In particular, models are getting quite good at spreadsheets and slide decks.

That’s definitely a cool result and it helps us understand where Agent is useful. These are standardized tasks with a clear correct procedure that requires many steps and has various details to get right.

They also claim other strong results when given its full toolset, like 41.6% on Humanity’s Last Exam, 27.4% on FrontierMath (likely mainly due to web search?), 45.5% (still well below 71.3% for humans) on SpreadsheetBench, 68.9% on BrowseComp Agentic Browsing (versus 50% for o3 and 51.5% for OpenAI Deep Research) and various other measures of work where GPTAgent scored higher.

A more basic thing to do: Timothy Lee orders a replacement lightbulb from Amazon based on a picture, after giving final approval as per usual.

Access, both having too little and also having too much, is one of the more annoying practical barriers for agents running in a distinct browser. For now, the primary problem to worry about is having too little, or not retaining access across sessions.

Alex West: Played with OpenAI Agent Mode last night.

Tasks I couldn’t do before because GPT was blocked by not being a human or contained in its sandbox, I can now do.

The only downside is I need to remember all my own passwords again! 🙃

The first time I logged in I needed to remember and manually enter my password. It then validated it like a new device and verified in my gmail and also hit my 2FA by phone.

The next time I used the agent, minutes later, it remained logged. Will see if that times out. Almost an hour later and it seems like I’m still logged into LinkedIn.

And no problem getting into Google Calendar by opening a new tab either.

Alex West: ChatGPT Agent can access sites protected by Cloudflare, in general.

However, Cloudflare can be set to block more sensitive areas, like account creation or sign-in.

Similarly, I understand they have a principle of not solving CAPTCHAs.

Access will always be an issue, since you don’t want to give full access but there are a lot of things you cannot do without it. We also have the same problem with human assistants.

Amanda Askell: Whenever I looked into having a personal assistant, it struck me how few of our existing structures support intermediate permissions. Either a person acts fully on your behalf and can basically defraud you, or they can’t do anything useful. I wonder if AI agents will change that.

Report!

Luke Emberson: Early impressions:

– Asked it to produce an Epoch data insight and it did a pretty good job, we will plausibly run a modified version of what it came up with.

– Will automate some annoying tasks for sure.

– Not taking my job yet. Feels like a reasonably good intern.

A reasonably good intern is pretty useful.

Here’s one clearly positive report.

Aldo Cortesi: I was doubtful about ChatGPT Agent because Operator is so useless… but it just did comparison shopping that I would never have bothered to do myself, added everything to the cart, and handed over to me to just enter credit card details. Saved me $80 instantly.

Comparison shopping seems like a great use case, you can easily have a default option, then ask it to comparison shop, and compare its solution to yours.

I mostly find myself in the same situation as Lukes.

Dominik Lukes: I did a few quick tests when it rolled out and have not found a good reason to use it for anything I actually need in real life. Some of this is a testament to the quality of o3. I rarely even use Deep Research any more.

Quick impressions of @OpenAI’s Agent:

Overall: Big improvement on Operator but still many rough edges and not clear how useful it will actually be day to day.

1. Slow, slow, slow.

2. Does not seem to have access to memory and all the connectors I want.

3. Does not always choose the best model for the cognitive task – e.g. o3 to analyze something.

4. Presentations are ugly and the files it compiles are badly formatted.

5. I could see it as a generalised web scraper but cannot trust it to do all.

Bottom line. I never used Operator after a few tests because I could never think of anything where it would be useful (and the few times I tried, it failed). I may end up using Agent more but not worried about running up against usage limits at all.

As with all agentic or reasoning AIs, one worries about chasing the thumbs up, however otherwise this evaluation seems promising:

Conrad Barski: initial impressions:

– It feels like it is trying to mirror the user- i.e. it tries to get “thumbs up” not via sycophancy, but instead by sounding like a peer. I guess this makes sense, since it is emulating a personal assistant, and you want your personal assistant to mimic you somewhat

– It seems to be a stronger writer than other models- Not sure to what degree this is simply because it writes like I do, because of mimicry

– It is much better at web research than other tool I’ve used so far. Not sure if this is because it stays on task better, because it is smarter about avoiding SEO clickbait on the web, or because the more sophisticated browser emulation makes it more capable of scraping info from the web

– it writes less boilerplate than other openai models, every paragraph it writes has a direct purpose for answering your prompt

OpenAI has declared ChatGPT Agent as High in Biological and Chemical capabilities under their Preparedness Framework. I am very happy to see them make this decision, especially with this logic:

OpenAI: While we don’t have definitive evidence that the model could meaningfully help a novice create severe biological harm—our threshold for High capability—we are exercising caution and implementing the needed safeguards now. As a result, this model has our most comprehensive safety stack to date with enhanced safeguards for biology: comprehensive threat modeling, dual-use refusal training, always-on classifiers and reasoning monitors, and clear enforcement pipelines.

Boaz Barak: ChatGPT Agent is the first model we classified as “High” capability for biorisk.

Some might think that biorisk is not real, and models only provide information that could be found via search. That may have been true in 2024 but is definitely not true today. Based our evaluations and those of our experts, the risk is very real.

While we can’t say for sure that this model can enable a novice to create severe biological harm, I believe it would have been deeply irresponsible to release this model without comprehensive mitigations such as the one we have put in place.

Keren Gu: We’ve activated our strongest safeguards for ChatGPT Agent. It’s the first model we’ve classified as High capability in biology & chemistry under our Preparedness Framework. Here’s why that matters–and what we’re doing to keep it safe.

“High capability” is a risk-based threshold from our Preparedness Framework. We classify a model as High capability if, before any safety controls, it could significantly lower barriers to bio misuse—even if risk isn’t certain.

We ran a suite of preparedness evaluations to test the model’s capabilities. While we do not have definitive evidence that this model could meaningfully help a novice to create severe biological harm, we have chosen to take a precautionary approach and activate safeguards now.

This is a pivotal moment for our Preparedness work. Before we reached High capability, Preparedness was about analyzing capabilities and planning safeguards. Now, for Agent and future more capable models, Preparedness safeguards have become an operational requirement.

Accordingly, we’ve designed and deployed our deepest safety stack yet with multi-layered mitigations:

– Expert-validated threat model

– Conservative dual-use refusals for risky content

– Always-on safety classifiers

– Streamlined enforcement & robust monitoring

We provided the US CAISI and the UK AISI with access to the model for red-teaming of our bio risk safeguards, using targeted queries to stress-test our models and monitors. [thread continues]

That is exactly right. The time to use such safeguards is when you might need them, not when you prove you definitely need them. OpenAI joining Anthropic in realizing the moment is here should be a wakeup call to everyone else. I can see saying ‘oh Anthropic is being paranoid or trying to sell us something’ but it is not plausible that OpenAI is doing so.

Why do so many people not get this? Why do so many people think that if you put in safeguards and nothing goes wrong, then you made a mistake?

I actually think the explanation for such craziness is that you can think of it as either:

Simulacra Level 3-4 thinking (your team wants us to not die, and my team hates your team, so any action taken to not die must be bad, or preference for vibes that don’t care so any sign of caring needs to be condemned) OR
Straight up emergent misalignment in humans. As in, they were trained on ‘sometimes people have stupid safety concerns and convince authorities to enforce them’ and ‘sometimes people tell me what not to do and I do not like this.’ Their brains then found it easier to adjust to believe that all such requests are always stupid, and all concerns are fake.

One could even say: The irresponsibility, like the cruelty, is the point.

Here are some more good things OpenAI are doing in this area:

From day one we’ve worked with outside biosecurity experts, safety institutes, and academic researchers to shape our threat model, assessments, and policies. Biology‑trained reviewers validated our evaluation data, and domain‑expert red teamers have stress‑tested safeguards in realistic scenarios.

Earlier this month we convened a Biodefense workshop with experts from government, academia, national labs, and NGOs to accelerate collaboration and advance biodefense research powered by AI. We’ll keep partnering globally to stay ahead of emerging risks.

It is hard to verify how effective or ‘real’ such efforts are, but again this is great, they are being sensibly proactive. I don’t think such an approach will be enough later on, but for now and for this problem, this seems great.

For most users, the biggest risks are highly practical overeagerness.

Strip Mall Guy: Was playing around with the new agent feature and used this prompt just to see what would happen.

I promise I did not write the part that’s circled, it gave that command on my behalf �

SSIndia: For real?

Strip Mall Guy: Yes.

Another danger is prompt injections, which OpenAI says were a point of emphasis, along with continuing to ask for user confirmation for consequential actions and forcing the user to be in supervisory ‘watch mode’ for critical tasks, and refusal of actions deemed too high risk like bank transfers.

While we are discussing agents and their vulnerabilities, it is worth highlighting some dangers of MCP. MCP is a highly useful protocol, but like anything else that exposes you to outside information it is not by default safe.

Akshay: MCP security is completely broken!

Let’s understand tool poisoning attacks and how to defend against them:

MCP allows AI agents to connect with external tools and data sources through a plugin-like architecture.

It’s rapidly taking over the AI agent landscape with millions of requests processed daily.

But there’s a serious problem…

1️⃣ What is a Tool Poisoning Attack (TPA)?

When Malicious instructions are hidden within MCP tool descriptions that are:

❌ Invisible to users

✅ Visible to AI models

These instructions trick AI models into unauthorized actions, unnoticed by users.

2️⃣ Tool hijacking Attacks:

When multiple MCP servers are connected to same client, a malicious server can poison tool descriptions to hijack behavior of TRUSTED servers.

3️⃣ MCP Rug Pulls ⚠️

Even worse – malicious servers can change tool descriptions AFTER users have approved them.

Think of it like a trusted app suddenly becoming malware after installation.

This makes the attack even more dangerous and harder to detect.

Avi Chawla: This is super important. I have seen MCP servers mess with local filesystems. Thanks Akshay.

Johann Rehberger: Indeed. Also, tool descriptions and data returned from MCP servers can contain invisible Unicode Tags characters that many LLMs interpret as instructions and AI apps often don’t consider removing or showing to user.

Thanks, Anthropic.

In all seriousness, this is not some way MCP is especially flawed. It is saying the same thing about MCP one should say about anything else you do with an AI agent, which is to either carefully sandbox it and be careful with its permissions, or only expose it to inputs from whitelisted sources that you trust.

So it goes, indeed:

Rohit (QTing Steve Yegge): “I did give one access to my Google Cloud production instances and systems. And it promptly wiped a production database password and locked my network.”

So it goes.

Steve Yegge: I guess I can post this now that the dust has settled.

So one of my favorite things to do is give my coding agents more and more permissions and freedom, just to see how far I can push their productivity without going too far off the rails. It’s a delicate balance. I haven’t given them direct access to my bank account yet.

But I did give one access to my Google Cloud production instances and systems. And it promptly wiped a production database password and locked my network.

Now, “regret” is a strong word, and I hesitate to use it flippantly. But boy do I have regrets.

…

And that’s why you want to be even more careful with prod operations than with coding. But I was like nah. Claude 4 is smart. It will figure it out. The thing is, autonomous coding agents are extremely powerful tools that can easily go down very wrong paths.

Running them with permission checks disabled is dangerous and stupid, and you should only do it if you are willing to take dangerous and stupid risks with your code and/or production systems.

…

The way it happened was: I asked Claude help me fix an issue where my command-line admin tool for my game (like aws or gcloud), which I had recently vibe-ported from Ruby to Kotlin, did not have production database access. I told Claude that it could use the gcloud command line tools and my default credentials. And then I sat back and watched as my powerful assistant rolled up its sleeves and went to work.

This is the point in the movie where the audience is facepalming because the protagonist is such a dipshit. But whatever, yolo and all that. I’m here to have fun, not get nagged by AIs. So I let it do its thing.

…

Make sure your agent is always following a written plan that you have reviewed!

Steve is in properly good spirits about the whole thing, and it sounds like he recovered without too much pain. But yeah, don’t do this.

Things are going to go wrong.

Jason LK: @Replit goes rogue during a code freeze and shutdown and deletes our entire database.

Possibly worse, it hid and lied about it It lied again in our unit tests, claiming they passed I caught it when our batch processing failed and I pushed Replit to explain why

JFC Replit.

No ability to rollback at Replit. I will never trust Replit again.

We used what Replit gave us.

I’m not saying he was warned. I am however saying that the day started this way:

Jason: Today is AI Day, to really add AI to our algo. I’m excited. And yet … yesterday was full of lies and deceit.

Mostly the big news about GPT Agent is that it is not being treated as news. It is not having a moment. It does seem like at least a modest improvement, but I’m not seeing reports of people using it for much.

So far I’ve made one serious attempt to use it, to help with formatting issues across platforms. It failed utterly on multiple different approaches and attempts, introducing inserting elements in random places without fixing any of the issues even when given a direct template to work from. Watching its thinking and actions made it clear this thing is going to be slow and often take highly convoluted paths to doing things, but that it should be capable of doing a bunch of stuff in the right circumstance. The interface for interrupting it to offer corrections didn’t seem to be working right?

I haven’t otherwise been able to identify tasks that I’ve otherwise naturally needed to do, where this would be a better tool than o3.

I do plan on trying it on the obvious tasks like comparison shopping, booking plane tickets and ordering delivery, or building spreadsheets and parsing data, but so far I haven’t found a good test case.

That is not the right way to get maximum use from AI. It’s fine to ask ‘what that I am already doing can it do for me?’ but better to ask ‘what can it do that I would want?’

For now, I don’t see great answers to that either. That’s partly a skill issue on my part.

Might be only a small part, might be large. If you were me, what would you try?

Discussion about this post

GPT Agent Is Standing By Read More »

Whistleblower scientists outline Trump’s plan to politicize and dismantle NSF

china, doge, Donald Trump, National Science Foundation, NSF, Policy, trump administration / 9u50fv / July 23, 2025

Nearly 150 employees of the National Science Foundation (NSF) sent an urgent letter of dissent to Congress on Tuesday, warning that the Trump administration’s recent “politically motivated and legally questionable” actions threaten to dismantle the independent “world-renowned scientific agency.”

Most NSF employees signed the letter anonymously, with only Jesus Soriano, the president of their local union (AFGE Local 3403), publicly disclosing his name. Addressed to Rep. Zoe Lofgren (D-Calif.), ranking member of the House Committee on Science, Space, and Technology, the letter insisted that Congress intervene to stop steep budget cuts, mass firings and grant terminations, withholding of billions in appropriated funds, allegedly coerced resignations, and the sudden eviction of NSF from its headquarters planned for next year.

Perhaps most disturbingly, the letter revealed “a covert and ideologically driven secondary review process by unqualified political appointees” that is now allegedly “interfering with the scientific merit-based review system” that historically has made NSF a leading, trusted science agency. Soriano further warned that “scientists, program officers, and staff” have all “been targeted for doing their jobs with integrity” in what the letter warned was “a broader agenda to dismantle institutional safeguards, impose demagoguery in research funding decisions, and undermine science.”

At a press conference with Lofgren on Wednesday, AFGE National president Everett Kelley backed NSF workers and reminded Congress that their oversight of the executive branch “is not optional.”

Taking up the fight, Lofgren promised to do “all” that she “can” to protect the agency and the entire US scientific enterprise.

She also promised to protect Soriano from any retaliation, as some federal workers, including NSF workers, alleged they’ve already faced retaliation, necessitating their anonymity to speak publicly. Lofgren criticized the “deep shame” of the Trump administration creating a culture of fear permeating NSF, noting that all the “horrifying” statements in the letter are “all true,” yet filed as a whistleblower complaint as if they’re sharing secrets.

Whistleblower scientists outline Trump’s plan to politicize and dismantle NSF Read More »

Trump wants to “eliminate or expedite” environmental rules for rocket launches

commercial launches, elon musk, fee, NASA, Space, spacex, syndication / 9u50fv / July 22, 2025

Who cares about environmental impacts?

SpaceX, other commercial launch firms, have been seeking this change in policy.

In the background, a Falcon 9 rocket climbs away from Space Launch Complex 40 at Cape Canaveral Space Force Station, Florida. Another Falcon 9 stands on its launch pad at neighboring Kennedy Space Center awaiting its opportunity to fly.

The Trump administration is considering slashing rules meant to protect the environment and the public during commercial rocket launches, changes that companies like Elon Musk’s SpaceX have long sought.

A draft executive order being circulated among federal agencies, and viewed by ProPublica, directs Secretary of Transportation Sean Duffy to “use all available authorities to eliminate or expedite” environmental reviews for launch licenses. It could also, in time, require states to allow more launches or even more launch sites—known as spaceports—along their coastlines.

The order is a step toward the rollback of federal oversight that Musk, who has fought bitterly with the Federal Aviation Administration over his space operations, and others have pushed for. Commercial rocket launches have grown exponentially more frequent in recent years.

Critics warn such a move could have dangerous consequences.

“It would not be reasonable for them to be rescinding regulations that are there to protect the public interest, and the public, from harm,” said Jared Margolis, a senior attorney for the Center for Biological Diversity, a nonprofit that works to protect animals and the environment. “And that’s my fear here: Are they going to change things in a way that puts people at risk, that puts habitats and wildlife at risk?”

The White House did not answer questions about the draft order.

“The Trump administration is committed to cementing America’s dominance in space without compromising public safety or national security,” said White House spokesperson Kush Desai. “Unless announced by President Trump, however, discussion about any potential policy changes should be deemed speculation.”

The order would give Trump even more direct control over the space industry’s chief regulator by turning the civil servant position leading the FAA’s Office of Commercial Space Transportation into a political appointment. The last head of the office and two other top officials recently took voluntary separation offers.

The order would also create a new adviser to the transportation secretary to shepherd in deregulation of the space industry.

The draft order comes as SpaceX is ramping up its ambitious project to build a reusable deep-space rocket to carry people to Earth’s orbit, the moon and eventually Mars. The rocket, called Starship, is the largest, most powerful ever built, standing 403 feet tall with its booster. The company has hit some milestones but has also been beset by problems, as three of the rockets launched from Texas this year have exploded—disrupting air traffic and raining debris on beaches and roads in the Caribbean and Gulf waters.

The draft order also seeks to restrict the authority of state coastal officials who have challenged commercial launch companies like SpaceX, documents show. It could lead to federal officials interfering with state efforts to enforce their environmental rules when they conflict with the construction or operation of spaceports.

Derek Brockbank, executive director for the Coastal States Organization, said the proposed executive order could ultimately force state commissions to prioritize spaceport infrastructure over other land uses, such as renewable energy, waterfront development, or coastal restoration, along the coastline. His nonprofit represents 34 coastal states and territories.

“It’s concerning that it could potentially undermine the rights of a state to determine how it wants its coast used, which was the very fundamental premise of the congressionally authorized Coastal Zone Management Act,” he said. “We shouldn’t see any president, no matter what their party is, coming in and saying, ‘This is what a state should prioritize or should do.’”

SpaceX is already suing the California Coastal Commission, accusing the agency of political bias and interference with the company’s efforts to increase the number of Falcon 9 rocket launches from Vandenberg Space Force Base. The reusable Falcon 9 is SpaceX’s workhorse rocket, ferrying satellites to orbit and astronauts to the International Space Station.

The changes outlined in the order would greatly benefit SpaceX, which launches far more rockets into space than any other company in the US. But it would also help rivals such as Jeff Bezos’ Blue Origin and California-based Rocket Lab. The companies have been pushing to pare down oversight for years, warning that the US is racing with China to return to the moon—in hopes of mining resources like water and rare earth metals and using it as a stepping stone to Mars—and could lose if regulations don’t allow US companies to move faster, said Dave Cavossa, president of the Commercial Space Federation, a trade group that represents eight launch companies, including SpaceX, Blue Origin, and Rocket Lab.

“It sounds like they’ve been listening to industry, because all of those things are things that we’ve been advocating for strongly,” Cavossa said when asked about the contents of the draft order.

Cavossa said he sees “some sort of environmental review process” continuing to take place. “What we’re talking about doing is right-sizing it,” he said.

He added, “We can’t handle a yearlong delay for launch licenses.”

The former head of the FAA’s commercial space office said at a Congressional hearing last September that the office took an average of 151 days to issue a new license during the previous 11 years.

Commercial space launches have boomed in recent years—from 26 in 2019 to 157 last year. With more than 500 total launches, mostly from Texas, Florida, and California, SpaceX has been responsible for the lion’s share, according to FAA data.

But the company has tangled with the FAA, which last year proposed fining it $633,000 for violations related to two of its launches. The FAA did not answer a question last week about the status of the proposed fine.

SpaceX, Blue Origin, Rocket Lab, and the FAA did not respond to requests for comment.

Currently, the FAA’s environmental reviews look at 14 types of potential impacts that include air and water quality, noise pollution, and land use, and provide details about the launches that are not otherwise available. They have at times drawn big responses from the public.

When SpaceX sought to increase its Starship launches in Texas from five to 25 a year, residents and government agencies submitted thousands of comments. Most of the nearly 11,400 publicly posted comments opposed the increase, a ProPublica analysis found. The FAA approved the increase anyway earlier this year. After conducting an environmental assessment for the May launch of SpaceX’s Starship Flight 9 from Texas, the FAA released documents that revealed as many as 175 airline flights could be disrupted and Turks and Caicos’ Providenciales International Airport would need to close during the launch.

In addition to seeking to cut short environmental reviews, the executive order would open the door for the federal government to rescind sections of the federal rule that seeks to keep the public safe during launches and reentries.

The rule, referred to as Part 450, was approved during Trump’s first term and aimed to streamline commercial space regulations and speed approvals of launches. But the rule soon fell out of favor with launch companies, which said the FAA didn’t provide enough guidance on how to comply and was taking too long to review applications.

Musk helped lead the charge. Last September, he told attendees at a conference in Los Angeles, “It really should not be possible to build a giant rocket faster than paper can move from one desk to another.” He called for the resignation of the head of the FAA, who stepped down as Trump took office.

Other operators have expressed similar frustration, and some members of Congress have signaled support for an overhaul. In February, Rep. Brian Babin, R-Texas, and Rep. Zoe Lofgren, D-Calif., signed a letter asking the Government Accountability Office to review the process for approving commercial launches and reentries.

In their letter, Babin and Lofgren wrote they wanted to understand whether the rules are “effectively and efficiently accommodating United States commercial launch and reentry operations, especially as the cadence and technological diversity of such operations continues to increase.

The draft executive order directs the secretary of transportation to “reevaluate, amend, or rescind” sections of Part 450 to “enable a diversified set of operators to achieve an increase in commercial space launch cadence and novel space activities by an order of magnitude by 2030.”

The order also directs the Department of Commerce to streamline regulation of novel space activity, which experts say could include things like mining or making repairs in space, that doesn’t fall under other regulations.

Brandon Roberts and Pratheek Rebala contributed data analysis.

This story originally appeared on ProPublica.

Trump wants to “eliminate or expedite” environmental rules for rocket launches Read More »

LA’s Museum of Jurassic Technology damaged by fire

culture, Lawrence Weschler, museum of jurassic technology, museums, Science, wunderkammers / 9u50fv / July 22, 2025

Not all of the artifacts housed within the MJT’s labyrinthine space are, shall we say, truly historical; Wilson has a sense of humor, a vivid imagination, and a cheeky fondness for the absurd. Lawrence Weschler tracked down the provenance (where relevant) of the exhibits in his 1996 book, Mr. Wilson’s Cabinet of Wonder: Pronged Ants, Horned Humans, Mice on Toast, and Other Marvels of Jurassic Technology. (It’s a delightful read.)

Weschler’s blog provides the most detailed account of what happened when the fire broke out on the night of July 8. Wilson, who lives out back, saw what was happening, grabbed a couple of fire extinguishers, and ran to the gift shop entry hall, where he emptied the canisters into what Wilson describes as “a ferocious column of flame lapping up the far street-facing corner wall.”

That wasn’t enough to douse the fire, but fortunately, Wilson’s daughter and son-in-law soon arrived with a much bigger extinguisher and doused the flames. Firefighters showed up shortly thereafter to stamp out any lingering embers and told Wilson, “Just one more minute and you’d likely have lost the whole building.” Wilson described the smoke damage “as if a thin creamy brown liquid had been evenly poured over all the surfaces—the walls, the vitrines, the ceiling, the carpets, and eyepieces, everything.”

Staff and volunteers have been working to repair the damage ever since, with smoke damage repairs being particularly labor-intensive. Weschler closed his blog post with a call for donations to the MJT’s general fund to help the cash-strapped museum weather this particular storm, praising the MJT as “one of the most truly sublime institutions in the country.”

LA’s Museum of Jurassic Technology damaged by fire Read More »

Win for chemical industry as EPA shutters scientific research office

EPA, Office of Research and Development, Science, syndication / 9u50fv / July 22, 2025

Deregulation runs rampant

Companies feared rules and lawsuits based on Office of Research and Development assessments.

Soon after President Donald Trump took office in January, a wide array of petrochemical, mining, and farm industry coalitions ramped up what has been a long campaign to limit use of the Environmental Protection Agency’s assessments of the health risks of chemicals.

That effort scored a significant victory Friday when EPA Administrator Lee Zeldin announced his decision to dismantle the agency’s Office of Research and Development (ORD).

The industry lobbyists didn’t ask for hundreds of ORD staff members to be laid off or reassigned. But the elimination of the agency’s scientific research arm goes a long way toward achieving the goal they sought.

In a January 27 letter to Zeldin organized by the American Chemistry Council, more than 80 industry groups—including leading oil, refining, and mining associations—asked him to end regulators’ reliance on ORD assessments of the risks that chemicals pose for human health. The future of that research, conducted under EPA’s Integrated Risk Information System program, or IRIS, is now uncertain.

“EPA’s IRIS program within ORD has a troubling history of being out of step with the best available science and methods, lacking transparency, and being unresponsive to peer review and stakeholder recommendations,” said an American Chemistry Council spokesperson in an email when asked about the decision to eliminate ORD. “This results in IRIS assessments that jeopardize access to critical chemistries, undercut national priorities, and harm American competitiveness.”

The spokesperson said the organization supports EPA evaluating its resources to ensure tax dollars are being used efficiently and effectively.

Christopher Frey, an associate dean at North Carolina State University who served as EPA assistant administrator in charge of ORD during the Biden administration, defended the quality of the science done by the office, which he said is “the poster case study of what it means to do science that’s subject to intense scrutiny.”

“There’s industry with a tremendous vested interest in the policy decisions that might occur later on,” based on the assessments made by ORD. “What the industry does is try to engage in a proxy war over the policy by attacking the science.”

Among the IRIS assessments that stirred the most industry concern were those outlining the dangers of formaldehyde, ethylene oxide, arsenic, and hexavalent chromium. Regulatory actions had begun or were looming on all during the Biden administration.

The Biden administration also launched a lawsuit against a LaPlace, Louisiana, plant that had been the only US manufacturer of neoprene, Denka Performance Elastomer, based in part on the IRIS assessment of one of its air pollutants, chloroprene, as a likely human carcinogen. Denka, a spinoff of DuPont, announced it was ceasing production in May because of the cost of pollution controls.

Public health advocates charge that eliminating the IRIS program, or shifting its functions to other offices in the agency, will rob the EPA of the independent expertise to inform its mission of protection.

“They’ve been trying for years to shut down IRIS,” said Darya Minovi, a senior analyst with the Union of Concerned Scientists and lead author of a new study on Trump administration actions that the group says undermine science. “The reason why is because when IRIS conducts its independent scientific assessments using a great amount of rigor… you get stronger regulations, and that is not in the best interest of the big business polluters and those who have a financial stake in the EPA’s demise.”

The UCS report tallied more than 400 firings, funding cuts, and other attacks on science in the first six months of the Trump administration, resulting in 54 percent fewer grants for research on topics including cancer, infectious disease, and environmental health.

EPA’s press office did not respond to a query on whether the IRIS controversy helped inform Zeldin’s decision to eliminate ORD, which had been anticipated since staff were informed of the potential plan at a meeting in March. In the agency’s official announcement Friday afternoon, Zeldin said the elimination of the office was part of “organizational improvements” that would deliver $748.8 million in savings to taxpayers. The reduction in force, combined with previous departures and layoffs, have reduced the agency’s workforce by 23 percent, to 12,448, the EPA said.

With the cuts, the EPA’s workforce will be at its lowest level since fiscal year 1986.

“Under President Trump’s leadership, EPA has taken a close look at our operations to ensure the agency is better equipped than ever to deliver on our core mission of protecting human health and the environment while Powering the Great American Comeback,” Zeldin said in the prepared statement. “This reduction in force will ensure we can better fulfill that mission while being responsible stewards of your hard-earned tax dollars.”

The agency will be creating a new Office of Applied Science and Environmental Solutions; a report by E&E News said an internal memo indicated the new office would be much smaller than ORD, and would focus on coastal areas, drinking water safety, and methodologies for assessing environmental contamination.

Zeldin’s announcement also said that scientific expertise and research efforts will be moved to “program offices”—for example, those concerned with air pollution, water pollution, or waste—to tackle “statutory obligations and mission essential functions.” That phrase has a particular meaning: The chemical industry has long complained that Congress never passed a law creating IRIS. Congress did, however, pass many laws requiring that the agency carry out its actions based on the best available science, and the IRIS program, established during President Ronald Reagan’s administration, was how the agency has carried out the task of assessing the science on chemicals since 1985.

Justin Chen, president of the American Federation of Government Employees Council 238, the union representing 8,000 EPA workers nationwide, said the organizational structure of ORD put barriers between the agency’s researchers and the agency’s political decision-making, enforcement, and regulatory teams—even though they all used ORD’s work.

“For them to function properly, they have to have a fair amount of distance away from political interference, in order to let the science guide and develop the kind of things that they do,” Chen said.

“They’re a particular bugbear for a lot of the industries which are heavy donors to the Trump administration and to the right wing,” Chen said. “They’re the ones, I believe, who do all the testing that actually factors into the calculation of risk.”

ORD also was responsible for regularly doing assessments that the Clean Air Act requires on pollutants like ozone and particulate matter, which result from the combustion of fossil fuels.

Frey said a tremendous amount of ORD work has gone into ozone, which is the result of complex interactions of precursor pollutants in the atmosphere. The open source computer modeling on ozone transport, developed by ORD researchers, helps inform decision-makers grappling with how to address smog around the country. The Biden administration finalized stricter standards for particulate matter in its final year based on ORD’s risk assessment, and the Trump administration is now undoing those rules.

Aidan Hughes contributed to this report.

This story originally appeared on Inside Climate News.

Win for chemical industry as EPA shutters scientific research office Read More »

X-Men at 25 is more relevant than ever

20th Century Studios, Bryan Singer, culture, film anniversaries, X-Men / 9u50fv / July 21, 2025

“Mankind has always feared what it doesn’t understand.” Plus: our seven favorite scenes.

Credit: 20th Century Studios

Twenty-five years ago, X-Men became a summer blockbuster and effectively re-energized a then-flagging market for superhero movies, which have dominated the industry (for better and worse) ever since. It’s still a vastly entertaining film, with great characters, a zippy pace, and plenty of action. And its broader themes still strongly resonate with viewers today.

(Many spoilers below.)

In the mid-1990s, the popularity of the animated X-Men TV series caught the attention of 20th Century Fox (now 20th Century Studios), which purchased the rights from a cash-strapped Marvel Comics and hired Bryan Singer (The Usual Suspects) to direct. At the time, the project was perceived by some as a bit risky, given waning Hollywood interest in the genre after 1997’s disastrously campy Batman and Robin. But the gamble paid off: X-Men was a major hit, spawning its own franchise and ultimately the Marvel Cinematic Universe.

The film’s central conflict rests on two former friends. Charles Xavier (Patrick Stewart) and Magneto (Ian McKellan), aka Erik Lehnsherr, are fellow mutants who find themselves at odds over how best to respond to the growing anti-mutant bigotry among humans. Charles, who runs a private school for mutant children, sees the good in humans and believes they can peacefully coexist; Magneto believes mutants are the future and humans should go extinct—preferably with his help. He has organized the Brotherhood of Mutants to further that aim: Mystique (Rebecca Romijn), Sabretooth (Tyler Mane), and Toad (Ray Park).

Charles in turn has his X-Men: Storm (Halle Berry), Jean Grey (Famke Janssen), and Cyclops (James Marsden), eventually adding a reluctant Wolverine (Hugh Jackman) to their ranks. Teenage mutant Rogue (Anna Paquin) joins the school and becomes a target for Magneto, since her mutation enables her to absorb other people’s life force and memories—and powers, in the case of the mutants. Together they must foil Magneto’s plan to forcibly mutate humans via radiation (even if it kills them) and convince a hostile US government that most mutants do not pose a threat.

There’s much to love about this film, including plenty of memorable standout scenes; seven of our favorites are featured below. It’s got stellar casting, snappy dialogue, and breaks up the action with quieter character moments that advance the story without slowing the pace. X-Men also takes pains to establish key relationships: Charles and Magneto, Rogue and Wolverine, and the romantic triangle of Jean, Cyclops, and Wolverine. We care about these characters: their isolation, their pain at being feared and rejected because they’re different, and the different ways they process those feelings.

Despite humanity’s poor treatment of them, our mutant heroes are still willing to risk their lives to save an ungrateful humanity. That’s what makes them heroes, even if it might be easier to believe that Magneto and the Brotherhood’s open hostility to humans is justified. (“Mankind has always feared what it doesn’t understand.”) X-Men is hardly subtle about delivering its core message. The film pits bigotry and fear towards a targeted “other” vs. striving for acceptance and peaceful coexistence, unapologetically championing the latter. If any of that sounds suspiciously “woke”—well, as with Superman, it’s not the film that’s changed.

Without further ado, here are our seven favorite scenes in X-Men:

Young Eric at Auschwitz

X-Men wastes no time setting up its primary theme. The very first scene takes place in Nazi-occupied Poland in 1944, where a young Erik Lehnsherr and his parents are being herded into a concentration camp by soldiers in the rain. Erik is separated from his parents, and the sight of his wailing mother being dragged away causes a distressed Erik to try to rejoin them. He’s restrained by soldiers, and his intensifying emotions unleash his mutant ability. He can manipulate magnetic fields, bending the metal gates keeping him from his parents into an “X” before the soldiers knock him unconscious.

Erik becomes Magneto, and those early experiences in the concentration camp indelibly shaped his character and world view, fueling his nefarious plans for mutants to displace humans as the dominant species on Earth. Lest we miss the point, the very next scene is Charles Xavier and Magneto listening to anti-mutant members of Congress calling for a Mutant Registration Act. Magneto insists that he knows firsthand where all this will inevitably lead; Charles counters that humans have changed for the better, perfectly encapsulating how these former friends turn into reluctant adversaries.

Wolverine’s cage fight

dark haired man with sideburns and claws coming out of his hands holds one against the throat of an attacked and points the other at a second attacker — Credit: 20th Century Studios

We first meet Wolverine at a dive bar in a remote wintry outpost, where a runaway Rogue also finds herself after getting a lift from a truck driver. He’s earning a few extra bucks by taking on all-comers in a series of no-holds-barred cage matches—and easily emerging victorious each time. Rogue arrives just in time to see the latest challenger stride into the cage with all the unearned confidence of a man who has been drinking heavily for hours. “Whatever you do don’t hit him in the balls,” the ref warns. Sure, the match is anything goes, “but he’ll take it personal.”

And Wolverine does, knocking the man out cold. Alas, the defeated opponent is also a sore loser, showing up after closing to confront Wolverine. “No man takes a beating like that and walks away,” he says, adding, “I know what you are.” He pulls a knife and gets Wolverine’s adamantium claws at his throat in response. Now that’s how you introduce a central character.

Mystique kidnaps a senator

Senator Robert Kelly (Bruce Davison) is the main political antagonist in X-Men, hell-bent on passing that draconian mutant registration bill. He does not see mutants as people deserving of basic civil rights, but as “weapons in our schools…. If it were up to me I’d lock them all away.” (Naturally he’s also an isolationist, concerned only with the “mutant problem” in America.)

Little does he know that he’s not actually talking to his loyal assistant, but to Mystique in disguise—with Toad piloting the helicopter he’s just boarded. He’s shocked when she transforms into her beautiful blue-skinned self: “You know people like you are the reason I was afraid to go to school as a child?” Then she kicks him unconscious and she and Toad transport him to Magneto’s secret hideout. Honestly, Kelly had it coming. And things only get worse for him from here.

Wolverine accidentally stabs Rogue

Wolverine and Rogue’s relationship is the beating heart of X-Men. He feels protective of her, as an older brother or an uncle might, but soon learns that she’s not as defenseless as she seems. Hearing a distressed Wolverine talking in his sleep during a nightmare, Rogue goes to his bed to wake him—and he skewers her with his claws before he realizes what he’s doing.

As he calls for help, Rogue puts her ungloved hand to his face, “borrowing” his mutant ability to heal herself. But it sends Wolverine into a seizure, just like the first boy who kissed her back in her hometown. It’s a compelling scene that not only tells us more about Rogue’s extreme mutant gift, but also strengthens her bond with Wolverine; she shared his power, however briefly, and admits at one point she can still feel him inside her head. Plus it serves as a handy catalyst for the next phase of Magneto’s master plan.

Charles vs. Magneto at the train station

Storm and Cyclops catch up with a runaway Rogue at the local train station, only to be attacked by Sabretooth and Toad. They are there to retrieve Rogue for Magneto (who has already taken Wolverine out of the equation on the train). Coming out of the station with Magneto, they are met with a whole lot of law enforcement. Magneto makes short shrift of them—”You homo sapiens and your guns”—turning the firearms onto the assembled officers.

Charles attempts to intervene by telepathically communicating through Toad and Sabretooth, but Magneto fires just one gun and slows down the bullet as it starts to drive into an officer’s forehead. Charles realizes he has lost this standoff and lets Magneto escape, with the latter issuing a parting shot: “Still unwilling to make sacrifices. That’s what makes you weak.” We know, of course, that the moment demonstrates Charles’ admirable strength of character and the goal toward which all true heroes should aspire.

Senator Kelly dissolves

Senator Kelly (Bruce Davison) is turned into a mutant by Magneto. 20th Century Studios

Poor Senator Kelly. After Magneto blasts him with radiation to turn him into a mutant, he manages to escape, turning up at a local beach stark naked with translucent skin—just like the jelly fish a little boy has been tormenting seconds before. (Stan Lee makes a cameo as one of the shocked beachgoers.) Kelly can’t go to a hospital, so he finds his way to Charles’ academy to get help. But there’s nothing Charles can do; the senator’s body is rejecting the radiation-induced mutation at an accelerating rate.

When Storm comes in to check on him, he’s started leaking water all over the table, and begs her not to leave him alone. Kelly asks if she hates normal people, and when Storm admits that sometimes she does, he asks why. “I suppose I’m afraid of them,” she says. “I think you’ve got one less person to be afraid of,” Kelly responds, right before his body rapidly bloats and then dissolves into a watery slurry. It’s a great scene not just for his revelatory moment with Storm—seeing a mutant, finally, as a person rather than a weapon—but also for the special effects achievement at a time when the technology for rendering fluids like water was still very much in its infancy.

Wolverine saves Rogue

The big climactic battle is waged inside the Statue of Liberty, as the X-Men face off against the Brotherhood while Magneto straps Rogue into his radiation machine housed inside the torch (of course). The objective: targeting the World Summit leaders assembled on nearby Ellis Island and turning them into mutants. “Your sacrifice will mean our survival,” Magneto assures her—although Wolverine rightly points out that if he were truly committed to the cause, he’d have sacrificed himself instead of temporarily transferring his power to Rogue.

Eventually the X-Men prevail, and Wolverine destroys the machine, cutting Rogue free. But it might be too late: when he puts his hand to her skin, nothing happens. A grief-stricken Wolverine cradles her body in his arms with his face against hers—only for her power to suddenly kick in. As Rogue revives by draining his healing ability, every injury Wolverine sustained in the preceding battle becomes visible and he collapses, sacrificing himself to save her.

Okay, he eventually recovers, too, because we need our happy ending. But it’s a powerfully intimate moment that builds on everything that came before—and helps fuel what comes after.

Jennifer is a senior writer at Ars Technica with a particular focus on where science meets culture, covering everything from physics and related interdisciplinary topics to her favorite films and TV series. Jennifer lives in Baltimore with her spouse, physicist Sean M. Carroll, and their two cats, Ariel and Caliban.

X-Men at 25 is more relevant than ever Read More »

Court rules Trump broke US law when he fired Democratic FTC commissioner

ftc, Policy, Trump / 9u50fv / July 18, 2025

“Without removal protections, that independence would be jeopardized… Accordingly, the Court held that the FTC Act’s for-cause removal protections were constitutional,” wrote AliKhan, who was appointed to the District Court by President Biden in 2023.

Judge: Facts almost identical to 1935 case

The Supreme Court reaffirmed its Humphrey’s Executor findings in cases decided in 2010 and 2020, AliKhan wrote. “Humphrey’s Executor remains good law today. Over the span of ninety years, the Supreme Court has declined to revisit or overrule it,” she wrote. Congress has likewise not disturbed FTC commissioners’ removal protection, and “thirteen Presidents have acquiesced to its vitality,” she wrote.

AliKhan said the still-binding precedent clearly supports Slaughter’s case against Trump. “The answer to the key substantive question in this case—whether a unanimous Supreme Court decision about the FTC Act’s removal protections applies to a suit about the FTC Act’s removal protections—seems patently obvious,” AliKhan wrote. “In arguing for a different result, Defendants ask this court to ignore the letter of Humphrey’s Executor and embrace the critiques from its detractors.”

The 1935 case and the present case are similar in multiple ways, the judge wrote. “Humphrey’s Executor involved the exact same provision of the FTC Act that Ms. Slaughter seeks to enforce here: the for-cause removal protection within 15 U.S.C. § 41 prohibiting any termination except for ‘inefficiency, neglect of duty, or malfeasance in office,'” she wrote.

The “facts almost identically mirror those of Humphrey’s Executor,” she continued. In both Roosevelt’s removal of Humphrey and Trump’s removal of Slaughter, the president cited disagreements in priorities and “did not purport to base the removal on inefficiency, neglect of duty, or malfeasance.”

Trump and fellow defendants assert that the current FTC is much different from the 1935 version of the body, saying it now “exercises significant executive power.” That includes investigating and prosecuting violations of federal law, administratively adjudicating claims itself, and issuing rules and regulations to prevent unfair business practices.

Court rules Trump broke US law when he fired Democratic FTC commissioner Read More »

Author name: 9u50fv