Connect with us

Tech

Plastics Companies Are Writing Lesson Plans. What Could Go Wrong?

Published

on

This story originally appeared on Grist and is part of the Climate Desk collaboration.

This fall, 54 million K-12 students are headed back to the classroom for another year of lessons in all the classic subjects: math, English, history, biology. But there’s another subject that’s been sneaking into school curricula: plastics.

A new report from the nonprofit Plastic Pollution Coalition documents the many ways the plastics industry has been inserting its agenda into classrooms across the US—including through lesson plans and worksheets, hands-on science experiments, and a program called “PlastiVan” that travels from city to city teaching students about “the contribution plastics make to modern life.”

Industry interests offer these resources to teachers for free or for cheap, according to the report. The materials tend to highlight the necessity of plastics while downplaying their significant downsides to human health and the environment.

Advertisement

“When that information is provided to kids, it obfuscates the true impacts of plastic pollution,” said Madison Dennis, senior policy and advocacy manager for Plastic Pollution Coalition and one of the report’s main authors. Besides harming marine life, plastics can expose people to hazardous chemicals, clog storm drains and contribute to flooding, and release planet-warming greenhouse gases. The Plastic Pollution Coalition is calling for stricter school policies against the use of industry-sponsored learning materials.

The report lays out four case studies of plastics industry “propaganda” for schoolkids. One involves the Society of Plastics Engineers, or SPE, a trade group that recently became a division of the Plastics Industry Association. SPE’s lesson plans include activities like a “plastic scavenger hunt,” which aims for students to “understand that plastics are ubiquitous and their importance to society and their personal life.” A video titled “Your Bottle Means Jobs” explains how plastics recycling supports local employment.

Another case study highlights the American Association of Chemistry Teachers, an initiative of the Dow Chemical Company. One of the association’s lesson plans teaches elementary and middle school students how to compare the strength of various types of plastic bags that can be found in a grocery store. After a brief experiment, the lesson invites them to “pretend [they] are an employee for Dow Chemical Company” and are designing a plastic bag for a customer.

While many of the materials claim to support STEM learning objectives, they do so while promoting familiar industry talking points, emphasizing the affordability and safety of plastics. Some acknowledge plastic pollution as a serious problem, but they blame irresponsible consumer behaviors and, instead of recommending less plastic be produced, propose more recycling as the primary way to address it.

Advertisement

In reality, only 9 percent of all plastics are recycled worldwide, and scientists have repeatedly warned that recycling will be unable to keep up with projected growth in plastic production. Investigative reporting has shown that industry groups knew this decades ago, but promoted recycling anyway in order to defuse growing concern over plastic pollution.

Wendy Johnson, a science education specialist at the National Center for Science Education, said the industry’s lesson plans and worksheets don’t reflect good pedagogy. Materials claiming to be aligned with state-level STEM standards don’t say which standards they support, or how they’re aligned. Lesson plans instruct students to read teachers’ notes, or list industry-specific vocab terms like “stretch blow molding” or “thermosetting.”

“By using all of these technical terms, it’s tricking people into thinking … that they’re doing science,” Johnson told Grist. What the materials really teach, she said, is “ideological perspectives about the economy,” like the desirability of cheap consumer goods.

Source link

Advertisement
Continue Reading
Click to comment

You must be logged in to post a comment Login

Leave a Reply

Tech

Why You Should Think Twice Before Buying That High-End Wi-Fi 7 Router

Published

on





Since arriving on shelves in 2023, Wi-Fi 7 routers have enticed consumers with promises of lightning-fast internet access and other benefits. However, while Wi-Fi 7 delivers on many of its selling points, that doesn’t necessarily mean you need to spend extra money on a premium Wi-Fi 7 router. Even if you decide to buy one of these routers, a mid-range model is often all you need for most households. That said, you may not even be able to take advantage of the benefits Wi-Fi 7 offers just yet.

It’s true that Wi-Fi 7 routers are an upgrade over their previous iterations. Although results vary based on numerous factors, they can offer throughput speeds up to about three times faster than their Wi-Fi 6 equivalents. Tests also reveal that Wi-Fi 7 routers tend to beat routers of previous generations when it comes to maintaining signal strength. CNET‘s testing reveals that the 15 Wi-Fi routers in its lab averaged 515 Mbps at 50 feet across all three bands, outperforming the Wi-Fi 6 routers by a significant margin, as those averaged 383 Mbps. The 6GHz band plays a big part here, though that band isn’t new to Wi-Fi 7.

Those benefits are all real, so if you’re considering upgrading to a Wi-Fi 7 router, that makes sense. You just need to make sure you’re not breaking the bank when doing so, as premium models aren’t worth it for most people.

Advertisement

Things to consider when buying a Wi-Fi 7 router

First, it’s worth noting that not all devices support Wi-Fi 7. That doesn’t mean a device that doesn’t support the technology won’t be able to connect to the internet via a Wi-Fi 7 router, but it does mean that it may be unable to take full advantage of the tech.

This factor alone shouldn’t make or break your decision to get a Wi-Fi 7 router. Even if your current devices don’t fully support the technology, newer, upgraded ones likely will. However, if you’re hoping for instant improvements across every piece of tech you own, that’s probably not going to happen.

Advertisement

Not all Wi-Fi 7 routers are created equal. Some Wi-Fi 7 routers are tri-band, while others are only dual-band, which means they don’t offer the fastest 6GHz option. Tests have shown that only the tri-band models are the better buy, as they’re the ones that offer a true performance uplift.

Lastly, be aware that your internet plan is most likely the ceiling, no matter how capable the router. The Federal Communications Commission (FCC) tested the advertised download speed across several ISPs, and the weighted average came out to 467 Mbps. You might not need a router that exceeds what your plan offers. On top of that, the 6 GHz band that Wi-Fi 7 offers may be fast if you’re near the router, but since it can’t penetrate walls and other obstacles as well as other frequencies, it might not properly serve your household.

Advertisement

You might not need the expensive Wi-Fi 7 router

Ultimately, a Wi-Fi 7 router isn’t necessary for everyone, and the same is true about premium models. You should assess your needs if you’re not sure whether making the upgrade is worthwhile. If you live in a household where the Wi-Fi connection needs to support multiple devices, gaming, video calls, large file downloads, etc., you’re the target consumer for this technology. If your Wi-Fi needs are fairly standard, you can probably wait to make the switch.

If you do decide to go ahead with buying a Wi-Fi 7 router, it’s still important to understand that you might not need to spend hundreds of dollars on a high-end model. You’re likely in the market for one of these routers because speed is a priority. Fortunately, tests indicate that you can find Wi-Fi 7 routers that deliver on this front for around $200 or less. Similarly, rankings of Wi-Fi 7 routers on websites like RTINGS indicate that, while the top-performing models may technically be the premium ones, the mid-range units tend to be more than good enough for the average household.

If you’re still unsure whether Wi-Fi 7 is right for you, you might want to do more research into the features that differentiate Wi-Fi 7 from Wi-Fi 6 before making a decision. It may also be necessary to learn about how to upgrade your devices so they’re more compatible with Wi-Fi 7.

Advertisement



Source link

Advertisement
Continue Reading

Tech

Peers ask why UK cyber bill leaves execs off the personal liability hook

Published

on

Peers have questioned why the UK’s Cyber Security and Resilience Bill does not allow regulators to penalize senior executives when an organization’s failure to comply involves their consent, connivance, or deliberate or careless neglect.

Echoing arguments heard across the industry for years, Baronesses Kidron and Ludford backed probing amendments that would introduce personal civil liability for senior execs and make cybersecurity a board-level responsibility.

“The intention behind the amendment is to change the culture of an organization, to ensure preventative action is taken, to avoid penalties,” said Baroness Kidron. “As I said at the outset, culture change starts at the top.”

Advertisement

The Register has previously reported on calls for NHS organizations, some of which would be covered by the bill’s reforms, to treat cybersecurity as a board-level priority.

More recently, 60 organizations committed to the aims of the UK government’s Cyber Resilience Pledge, promising to ensure their boards take responsibility for their organization’s cybersecurity.

Peers supporting the amendments pointed to financial sector rules introduced over the past decade that can impose regulatory or criminal liability on the C-suite for serious failings.

They argued that the amendments would bring the bill closer to the EU’s NIS2 directive, which includes senior management accountability measures. Personal liability is not mandatory under NIS2, however, and member states have implemented it differently.

Advertisement

Supporting the personal liability proposal, Lord Clement-Jones said: “If an individual is fit to draw a multimillion-pound executive salary running a critical national provider, they must be prepared to carry personal responsibility for securing it.”

Despite support from several peers, the government defended its existing plan to impose substantial maximum fines and introduce security, resilience, and governance requirements through secondary legislation.

“It is absolutely right that organizations, especially those delivering our essential services, are held properly accountable for their activities,” said cybersecurity minister Baroness Lloyd of Effra, who did not support the personal liability amendment.

She cited the maximum fines of £17 million or 4 percent of the offending organization’s annual turnover, whichever is higher, calling it “a meaningful enforcement regime.”

Advertisement

Baroness Lloyd said the forthcoming security and resilience requirements would mandate board-level governance in line with the NCSC’s Cyber Assessment Framework. The government has yet to consult on the details.

“That will cover issues such as organisational capability, senior responsibility, and accountability for security and resilience and effective risk escalation, and it is in that way that we will connect the clarity on what is expected of boards with the accountability through the enforcement regime.”

Reporting requirements and other matters

Separately, peers quizzed the government on the structure of the bill’s strict reporting requirements, warning that the current wording threatens to overwhelm regulators with an administrative burden.

One of the CSR bill’s primary objectives is to collect more data about the threats facing UK organizations by imposing stricter reporting requirements on in-scope entities. 

Advertisement

The bill requires regulated organizations to issue an initial notification within 24 hours and a fuller report within 72 hours. It defines an incident as an event that has, or is capable of having, an adverse effect on an operation.

Former security minister Baroness Neville-Jones suggested changing the wording from “capable of” to “likely to have,” to reduce the reporting burden on regulated organizations.

Lord Clement-Jones agreed, warning that the current wording would “unleash an administrative tsunami of defensive reporting.”

He also argued that the government’s definition of a data compromise was overly broad and “dramatically expanding the notification net to include technical data anomalies that cause zero disruption or loss to actual customers.”

Advertisement

He said the reporting rules and broad definition of compromise could leave responsible operators of essential services spending more time on paperwork than improving their defenses.

The government was unmoved, and Baroness Lloyd said that the more stringent reports were crucial in achieving the aims of the bill, which seeks to update the existing NIS Regulations 2018.

While some peers were looking to ease the burden of reporting, others sought to increase it in other areas.

Baroness Harding, who is uniquely placed to weigh in on cyberattack response, proposed a 14-day intermediate report and a final report due one month after the attack first occurred.

Advertisement

Drawing on her experience as the former TalkTalk CEO, she said that after 72 hours, attacked organizations start to get “real data,” but “it’s really only after a couple of weeks that you’ve got a proper sense of what has happened.”

The final report comes at the one-month mark, when “the fog is starting to clear and you have a proper sense of the real scale of the problem,” she said.

Harding said that senior executives are typically told from all sides not to say anything about a cyberattack, but this only serves the criminals, who meanwhile may be attacking other victims.

“If you share this information in the fog with regulators and with law enforcement agencies, that’s how the law can prevail,” she said.

Advertisement

“It’s how regulators can work out what’s happening, it’s how they can warn others who might be affected, and it’s how the law enforcement agencies can do their work to actually try and find the bad guys.”

Baroness Lloyd defended the bill’s existing two-stage process, arguing that it already provides information at the points when regulators need it.

She reaffirmed that the initial 24-hour report alerts the NCSC and allows it to determine whether other organizations are affected, while the 72-hour report will contain the necessary details to enable a more actionable response.

“We believe that the stages we set out meet that. They have been carefully developed to provide the right notification at the appropriate time,” she told peers.

Advertisement

“They have been developed in consultation with industry, as many noble Lords exhorted in the previous group.

“Crucially, under the information-gathering powers in Clause 15, regulators can also request further information about an incident that has been reported to them if they consider this necessary to understanding the incident and how it is being managed.

“Obviously, that may be appropriate in some incidents and not in others. That kind of practical balance is enabled by the bill.”

Separately, the Grand Committee spent the second day scrutinizing the bill, fleshing out datacenters’ responsibilities, as well as examining requirements to notify affected downstream customers within 24 hours of a breach instead of 72 hours.

Advertisement

The government also rejected concerns that cybersecurity data collected under the reporting rules could contribute to unfair overseas proceedings. Baroness Lloyd said ministers had considered the issue and assessed the risk as very low. ®

Source link

Continue Reading

Tech

Comparing PETG And PCTG Filaments

Published

on

The average 3D printer owner knows a few types of filaments – PLA, ABS, somewhere in the middle, PETG.  PCTG is another option that can be confusingly similar to PETG. Recently, [Igor Gaspar] of [My Tech Fun] took a poke at both types. He obtained both PETG and PCTG transparent filaments from the same manufacturer to compare them directly.

As we recently detailed in an article on PET polyesters, PETG is glycol-modified PET, meaning that some of the glycol monomers are replaced by CHDM monomers to create a more flexible and robust material. PCTG is very similar to PETG, except that more than half of the glycol monomers are replaced rather than less than half. This creates a PET-type material that has distinct physical properties from PETG, which might be desirable for some applications.

PCTG is more ductile due to the addition of more CHDM, but also requires higher temperatures to print, closer to ASA presets. During testing, it’s obvious that PCTG is indeed much more flexible, making it potentially a good choice for springs and compliant mechanisms. PCTG is also highly impact-resistant, unlike PETG, and resists higher temperatures much better.

Advertisement

Overall, other than the higher printing temperatures, PCTG seems like a solid option for more extreme environments, potentially as an alternative to ASA and similar filaments.

Advertisement

Source link

Continue Reading

Tech

Startup takes on AI hallucinations with $25M and an HQ rooted in a small Washington town

Published

on

Kevin Owens, co-founder and CEO of Resect AI. (Resect AI Photo)

Resect AI, an artificial intelligence startup led by a team of scientists and engineers in Washougal, Wash., launched out of stealth Thursday with $25 million in funding to commercialize an open-source technology designed to catch AI hallucinations before they happen.

Unlike traditional AI monitoring tools that evaluate generated text after the fact, Resect AI says its patented technology operates in-stream — looking deep inside large language models in real time to observe, detect, interpret, and modify model behavior before a hallucination can occur.

By intervening directly within the model’s internal decision-making process rather than running post-hoc checks, the platform stops fabrications at the source while simultaneously generating an audit trail for enterprise compliance and due diligence.

“AI has prematurely been put in a position of trust. Adding labels such as ‘use at your own risk’ flies in the face of proper governance or compliance,” Kevin Owens, co-founder and CEO of Resect AI, said in a news release. “We are building the next large enterprise AI company to bring transparency and accountability to AI for industries such as publishing, finance, healthcare, research, and education where factual accuracy is absolutely critical.”

Beyond its tech, the startup’s leadership is also bullish about its small-town presence.

Advertisement

Washougal is a city of roughly 18,000 residents, 175 miles south of Seattle, tucked along the Columbia River across from Portland. Resect AI employs four people at an office on Main Street — including its co-founders — out of a 30-person workforce spread across the Seattle area, California, New York, and Texas.

“We believe the talent is up to par and we loved the sense of community that we found when we first came up here,” Owens told GeekWire. “We have been coming to the greater Washington and Oregon areas on and off over the years and finally decided this needed to be our headquarters.”

Owens said the decision has already paid off, noting that the startup has quickly tapped into the region’s talent pool by recruiting PhDs from both the greater Seattle and Portland markets while connecting with Northwest capital markets leaders.

Resect AI is also planning to open an office in the Seattle area in the near future for engineering and to serve as a business hub.

Advertisement

Alongside Owens, Resect’s other co-founders include Tim Walton, chief artificial intelligence officer; Tyler Gerber, chief operating officer; and Tommy Lofgren, chief product and marketing officer.

The company plans to use the funding to accelerate research and development, expand its go-to-market initiatives, and fuel talent acquisition — bringing its total headcount to 50 by the end of 2026.

Source link

Advertisement
Continue Reading

Tech

When Are Portable Apple CarPlay Screens Actually Worth It?

Published

on

If your vehicle is older, you might want to check this out.

CarPlay has become a must-have for many drivers. It’s no surprise, since it makes life so much easier when it comes to accessing your iPhone apps in the car. Navigation, music, messages — all right there on the dashboard. And while there are more than 800 CarPlay-compatible vehicles, you might have one that isn’t on the list. But no, you don’t need to trade in your car. Some portable screens can give you CarPlay for much less. They aren’t quite as good as built-in systems, but they get the job done.

Despite all the convenience CarPlay offers, some automakers have been getting rid of it. GM, for example, has decided to phase out CarPlay and Android Auto from its EVs. Yes, many people love CarPlay. But they want you to use their own systems. It’s an understandable business move, considering that some even charge you to unlock extra features in their cars. Getting a portable CarPlay screen ends up being a good way to get around this.

Advertisement

What to look for when buying a portable CarPlay screen

Unlike buying a new car, getting one of these portable CarPlay screens won’t cost you much. You can find many of them for under $50. Even if you pay $150 for one of these, it’s still much cheaper than a new car. Some have larger displays and will cost more, and some really cheap ones might be best to avoid — those usually have really bad touchscreens. Start by choosing the right size. Most of these range from 6 to 11 inches. Some users find large screens too distracting, so it might be worth going with a smaller one. Smaller screens are easier to fit in your car, too.

But size isn’t the only thing that matters here. There are other specs you should consider, like screen resolution. Some portable screens have very low resolution. It’s not as if any of these screens were designed for watching 4K video — they weren’t. But going with a low-resolution screen will make everything look terrible, from text to icons. Look for screens with a resolution of at least 1280×720 pixels. The audio output is also a big plus. Most of these screens have awful speakers, so be sure to choose one that supports AUX, Bluetooth or FM for playing audio. Also, make sure you’re getting wireless CarPlay instead of wired CarPlay. Not every portable screen lets you use CarPlay without a cable.

Advertisement

Extra features they can offer you

Even though these portable screens are designed for CarPlay, some of them offer interesting extra features. Many models include integrated dash cameras with microSD loop recording. Others might come with a rear camera, which is nice for adding parking assistance to your old car. Also, keep an eye on how the display attaches to your dash. Some of them have very unstable mounts, and you don’t want your screen flying off as you turn a corner. And keep an eye out for screens that come with Android built in. That doesn’t matter for running CarPlay, but it lets you install apps that can run without your iPhone.

As you can see, buying one of these portable screens is an easy way to get around the lack of CarPlay. They work and are cheaper than a retrofit to your car’s built-in system. Sure, they won’t look like a factory system, but that’s fine. For the price, you’re still getting a great deal.

Source link

Advertisement
Continue Reading

Tech

Banking giant UBS wants all new employees to have AI skills

Published

on


  • 2027 UBS recruits told they must have AI proficiency and a willingness to learn
  • The bank’s AI Fluency Pathway will continue to support junior workers’ development
  • While thousands of jobs could be at risk, we’re starting to see shifts rather than outright displacement

Swiss investment giant UBS is now requiring all junior bankers to demonstrate AI proficiency as the skill moves from being a nice-to-have to an absolute requirement within recruiting.

The change currently applies to graduates and interns applying to the company’s 2027 intake, per the Financial Times, and it’s unclear whether UBS will broaden the requirement to all workers in the future.

Source link

Continue Reading

Tech

Why AI visibility now demands paid and organic GEO optimization

Published

on

For the last two years, the advice to brands has been more or less standardized: get cited and recommended in AI, and the rest will follow.

Whether that is called GEO or AEO, the objective is the same: make sure AI can understand your brand, retrieve the right information and recommend it when a customer asks.

Source link

Advertisement
Continue Reading

Tech

US Military Disables Ad Trackers On Devices To Protect Troops

Published

on

Slashdot reader DeanonymizedCoward writes: Reuters reports that the US Military is disabling ad tracking on devices, to prevent adversaries from buying publicly-available tracking data to assist in targeting troops. Military officials say that they have disabled trackers on a variety of computers and mobile devices, according to letters released on Friday by Sen. Ron Wyden, and following reports that commercially-available tracking data has been used to target troops in the Middle East….

Wyden said in a statement that it was clear that the military’s efforts “have not been effective at neutralizing this threat.” U.S. Representative Pat Harrigan, a North Carolina Republican, said that U.S. enemies “should not be able to pull out a credit card and buy information that helps them track American troops.” Rep. Harrigan remains silent as to the larger question of whether the general public should be able to pull out a credit card and buy information that helps them track anyone they please.

“The Pentagon said in an email it would respond to the lawmakers directly,” Reuters reports:

The Army said in a statement that advertising IDs had been blocked on Windows computers “since before 2021” but that Android and Apple mobile devices had only had it disabled by default “since at least February 2026….” The effort to reduce the location data generated by smartphones comes as military officials weigh increasingly strict restrictions on phone use overall. In July, Reuters reported that some deployed personnel in the Middle East could be ordered to surrender their phones amid concerns that mobile videos they were posting to the internet were helping Iran target American bases in the region.

Advertisement

Read more of this story at Slashdot.

Source link

Advertisement
Continue Reading

Tech

Roku's 24/7 AI Slop Channel Is Even Worse Than Expected

Published

on

A new streaming TV channel shows films made with genAI, reports Engadget. “Fairground AI Creator TV” is free — and supported with ads — describing its material as “AI Cinema”:

“AI-generated” can make it sound as though someone typed a sentence into a machine and came back five minutes later to find a finished movie. Fairground’s catalog shows why that description can be too simple. Take Lost Garden: The Awakening of the Lantern Knight. According to its Fairground page, creator Frank Houbre wrote the world, characters, mythology, emotional arc and screenplay himself. AI tools were used mainly for animation and visual production, with other tools helping create voices and music before the episode was assembled in conventional video-editing software.

There’s still one question, the article notes: “whether viewers actually want an AI-focused TV channel.” More than 100 AI creators have contributed to the 24-hour slate of programming, although Variety points out several of them were discovered on social media.

The channel was recently profiled in an article by the Guardian. Its headline? “‘Nightmare fodder’: Roku’s AI slop channel is even worse than expected.” (And its subheading calls it “a 24/7 channel devoted to low-quality AI content for viewers sick of watching real people move…”)
What about people who hate plot and vision and the sight of people speaking convincing dialogue that synchronises perfectly with the movement of their lips? What about the people who just want to watch an unyielding torrent of eerily weightless nightmare fodder? Well, good news. Roku has finally caught up… Early reactions were, to put it mildly, not great. The Verge compared it to eating from a trough, while Futurism called it “bottom-of-the-barrel slop”…

Advertisement

On the plus side, the channel is evidence that artificial intelligence has come on in leaps and bounds over the last couple of years… However, it is still awful. Categorically, catastrophically awful. The channel doesn’t so much offer shows as a drifting dreamscape of bad ideas rendered as horribly as possible with no thought paid to scheduling. At one point on Wednesday, a shrill high-frequency anime gave way to a long and staid German-language short about Nazi bureaucracy. After that came a sort of Gladiator ripoff that had all the dynamism of an exhibit you’d see at the fourth-best museum on a poorly planned family holiday.

Read more of this story at Slashdot.

Source link

Advertisement
Continue Reading

Tech

M4 Macs can share local AI work with PCs using Nvidia PAIR

Published

on

Got an M4 Mac? Nvidia’s new free tool lets it offload local AI tasks to your PC, turning your home network into a shared supercomputer.

The Personal AI Router (PAIR) isn’t hardware. Instead, it’s software that connects supported Apple Silicon Macs, Nvidia RTX PCs, and DGX Spark systems.

PAIR works with Ollama and LM Studio, two tools for running AI models locally. It discovers participating computers and gives AI apps a single connection for sending requests, so users don’t have to configure each app to reach every machine.

For someone already running local AI on a Mac, PAIR could put a compatible gaming PC to work when requests pile up. An AI agent reviewing several documents, for example, could have independent requests handled on different computers.

Advertisement

Nvidia released the open-source beta on September 3.

How PAIR distributes local AI requests

PAIR chooses an available computer for each request, while Ollama or LM Studio runs the model on that machine. When a computer is busy or unavailable, the router can direct new requests elsewhere.

The computers remain separate systems, according to Nvidia’s technical FAQ. PAIR doesn’t combine their GPUs or memory, so connecting two 16GB Macs doesn’t create one 32GB memory pool for a larger model.

Adding another computer doesn’t automatically make an individual AI response faster.

Advertisement

In a performance demonstration, Nvidia ran a task divided among five AI subagents in Hermes, using Ollama and the Qwen 3.6 35B A3B model. It reported average completion times of 18 minutes on an RTX Spark laptop and 8 minutes and 48 seconds across that laptop, a DGX Spark, and an RTX 5090.

Nvidia describes the result as an unofficial demonstration specific to that configuration. The comparison didn’t include a Mac, so it doesn’t establish how much a Mac user would gain from PAIR.

Mac requirements leave M3 Ultra outside the support list

Nvidia’s system requirements list macOS Tahoe and an M4 or newer chip for Mac support. The general requirements specify at least 8GB of RAM and recommend 20GB or more of disk space.

Dark webpage section titled What Can NVIDIA PAIR Can Do, featuring four white icons with short descriptions about powering AI clusters, moving workloads, using familiar tools, and keeping local inference closeFeatures of Nvidia PAIR

The cutoff leaves the M3 Ultra Mac Studio outside PAIR’s published hardware requirements. Apple introduced that Mac in March 2025 with configurations offering up to 512GB of unified memory, and its launch announcement specifically promoted running large language models locally.

Nvidia’s requirements page doesn’t explain the M4 cutoff, and the published support list doesn’t establish whether PAIR would work on an older Mac.

Advertisement

Supported Nvidia hardware includes GeForce RTX 20-series GPUs and newer, RTX Pro workstation GPUs based on the Turing architecture or newer, and DGX Spark systems. PAIR supports Windows and Linux alongside macOS.

Users need to install PAIR, pair their computers, and make the required models available through Ollama or LM Studio. Each participating machine still needs the resources to run the requested model.

Nvidia describes PAIR’s local inference as keeping prompts, files, and agent context on the user’s network. Models need to be downloaded first, but PAIR itself doesn’t require an internet connection to operate.

Advertisement

Source link

Continue Reading

Trending

Copyright © 2025