MAROKO133 Update ai: Researchers Just Found Something Extremely Alarming About AI’s Power

📌 MAROKO133 Hot ai: Researchers Just Found Something Extremely Alarming About AI’s

Researchers have found that the carbon footprint of generative AI-based tools that can turn text prompts into images and videos is far worse than we previously thought.

As detailed in a new paper, researchers from the open-source AI platform Hugging Face found that the energy demands of text-to-video generators quadruple when the length of a generated video doubles — indicating that the power required for increasingly sophisticated generations doesn’t scale linearly.

For instance, a six-second AI video clip consumes four times as much energy as a three-second clip.

“These findings highlight both the structural inefficiency of current video diffusion pipelines and the urgent need for efficiency-oriented design,” the researchers concluded in their paper.

Experts are warning that we’re rolling out generative AI tools without a full grasp of their true environmental impacts.

“Ultimately, we found that the common understanding of AI’s energy consumption is full of holes,” MIT Technology Review wrote in a recent analysis.

While image generators used the equivalent of five seconds of microwave warming to generate a single 1,024 x 1,024 pixel image, video generators proved far more energy-intensive. To spit out a five-second clip, the researchers found that it takes the equivalent of running a microwave for over an hour. If they’re consuming far more power as the length increases, the math doesn’t look good.

Those demands rise even faster for longer clips, implying “rapidly increasing hardware and environmental costs,” according to the Hugging Face researchers’ paper.

Fortunately, there are ways to slim down those demands, including intelligent caching, the reusing of existing AI generations, and “pruning,” meaning the sifting out of inefficient examples from training datasets.

But whether those efforts will be enough to make a dent in the enormous electricity consumption of current AI tools remains to be seen. The scale of its impact is substantial, with AI-related energy usage already representing 20 percent of global datacenter power demands, according to a recent study.

Meanwhile, tech giants are investing tens of billions of dollars in infrastructure buildouts, sometimes abandoning climate goals in the process. In its 2024 environmental impact report, Google admitted that it was woefully behind its ambitious plan to reach net-zero carbon emissions by 2030, seeing a staggering 13 percent increase in carbon emissions year over year, in large part due to its embrace of generative AI.

Earlier this year, the company released its Veo 3 AI video generator, later boasting that users had created over 40 million videos in just seven weeks.

While the environmental impact of the tool remains unknown — Google isn’t exactly incentivized to investigate its sizable contributions to carbon emissions — chances are it’s far worse than we think.

More on AI energy usage: How Much Electricity It Actually Takes to Use AI May Surprise You

The post Researchers Just Found Something Extremely Alarming About AI’s Power Usage appeared first on Futurism.

🔗 Sumber: futurism.com

📌 MAROKO133 Hot ai: Putting Mirrors on Traffic Cones Causes Self-Driving Cars to M

Researchers have engineered the ultimate nemesis of self-driving cars: mirror-adorned traffic cones.

In a series of tests highlighted by The Register, a team from France and Germany showed that their optical sleight of hand could easily dupe lidar-equipped autonomous cars into not recognizing obstacles on the road. 

Time after time, the experiments showed, the correct mirror placement left the cars oblivious and attempting to plow through sacrificial traffic cones. And if that was the disappearing act, they were also able to pull off a conjuring trick, deviously psyching out the car’s software into seeing obstacles that weren’t there.

“An adversary can inject phantom obstacles or erase real ones using only inexpensive mirrors,” the researchers warned in their new study, which is awaiting peer review. “These are practical threats capable of triggering critical safety failures, such as abrupt emergency braking and failure to yield.”

The experiments will add to the major safety concerns swirling around self-driving cars.

Lidar, short for light detection and ranging, uses rapid laser pulses to detect a car’s surroundings, similar to how sonar uses sound waves. Most autonomous vehicle companies rely on them — except for Tesla, whose CEO Elon Musk insists that the technology is an expensive “crutch.” (It’s worth noting, however, that his cameras only approach isn’t without its major flaws, like being blinded by sunlight.)

There’re already plenty of questions about self-driving cars’ ability to recognize everyday obstacles, including pedestrians. But there’s also concern how the tech will handle people deliberately trying to mess with it. Activists learned they could disable Waymo robotaxis by placing a traffic cone on their hoods — which in retrospect, is a perfect, dunce cap-esque symbol for the industry’s ongoing woes.

In the latest study, the researchers used cars powered by a popular open source software called Autoware. Making the cones disappear was fairly straight-forward: with enough tinkering, fully covering a traffic cone with a mirror was enough to turn them invisible to the lidar. The trick was to angle the mirror either towards the ground or the sky. This way, it deflects the laser pulses away from the obstacle. The researchers call this an ORA, or “object removal attack” — and it worked every single time.

The inverse, an “object addition attack” (OAA), requires more setup but can be just as effective. The researchers found that they could block a car from making a turn by putting a mirror on the corner of a sidewalk angled towards the vehicle. This causes the pulses to bounce off the vehicle’s body and create a “phantom echo” that, to the self-driving software, pops up along the path it’s turning into. And the more mirrors they used, the more the car’s software was convinced it was seeing an obstacle. With six mirrors, the lidar-dependent rides misclassified the phantom signals as a “CAR” with a 74 percent confidence.

This isn’t a deathblow for the tech, per se. But it is a big, fat reminder that there’s still some major limitations with the technology that warrants some — dare we say it — serious reflection.

More on self-driving cars: Tesla’s Robotaxis Have Already Gotten Into Numerous Accidents

The post Putting Mirrors on Traffic Cones Causes Self-Driving Cars to Melt Down, Confounding Lidar appeared first on Futurism.

🔗 Sumber: futurism.com