Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog – Page 1362

At the end of the day it just got too expensive to make games, and too risky to release bad ones. Not to mention the political nonsense. AI is now in the wings poised for a take over game development. Will of mostly taken over around 2030. And, it will quickly be back to the old days.

There’s one topic that’s stayed on my mind since the Game Developers Conference in March: generative AI. This year’s GDC wasn’t flooded with announcements that AI is being added to every game — unlike how the technology’s been touted in connection with phones and computers. But artificial intelligence definitely made a splash.

Enthusiasm for generative AI was uneven. Some developers were excited about its possibilities, while others were concerned over its potential for abuse in an industry with shattered morale about jobs and careers.

AI has been a common theme at GDC presentations in years past, but in 2024 it was clear that generative AI is coming for gaming, and some of the biggest companies are exploring ways to use it. With all new technologies, there’s no guarantee they’ll stick. Will generative AI flame out like blockchain and NFTs, or will it change the future of gaming?

Continue reading “Generative AI Is Coming for Video Games. Here’s How It Could Change Gaming” | >

Astronomers have picked up a gravitational-wave signal originating from a dramatic collision deep in the cosmos. The event, dubbed GW230529, was recorded by the LIGO Livingston detector in May 2023.

Gravitational waves are caused by the acceleration of massive objects, such as merging black holes or neutron stars. According to Albert Einstein’s theory of general relativity, massive objects like planets, stars, and black holes distort the fabric of spacetime around them.

When these massive objects accelerate or change speed, they create waves that propagate outward at the speed of light. The detection of gravitational waves opens up a new window for observing the universe, allowing scientists to study phenomena that were previously inaccessible, such as the mergers of black holes and neutron stars, as well as the nature of gravity itself.

Read more | >

Jones’ family home sat to the south of Lake Livingston, in the river bottoms of Coldspring, the San Jacinto County seat. It was overtaken by water shortly after the family left and Jones found safe harbor for their animals, his neighbors told him.

Much of the county was still underwater Friday as crews pulled stranded residents from their homes and roadways.

His family sat among dozens of evacuees who rested on cots and sat around plastic folding tables in Dunbar Gym, a makeshift shelter in an old school building. Many were elderly or infirm, few spoke English or were comfortable telling their stories.

Read more | >

Visual language models have evolved significantly recently. However, the existing technology typically only supports one single image. They cannot reason among multiple images, support in context learning or understand videos. Also, they don’t optimize for inference speed.

We developed VILA, a visual language model with a holistic pretraining, instruction tuning, and deployment pipeline that helps our NVIDIA clients succeed in their multi-modal products. VILA achieves SOTA performance both on image QA benchmarks and video QA benchmarks, having strong multi-image reasoning capabilities and in-context learning capabilities. It is also optimized for speed.

It uses 1 ⁄ 4 of the tokens compared to other VLMs and is quantized with 4-bit AWQ without losing accuracy. VILA has multiple sizes ranging from 40B, which can support the highest performance, to 3.5B, which can be deployed on edge devices such as NVIDIA Jetson Orin.

Read more | >

A new type of memory has been demonstrated running at an astounding 600C for over 60 hours. Non-volatile ferroelectric diode (ferrodiode) memory devices can offer outstanding heat resistance and other properties that should enable cutting-edge data and extreme environment computing, claim researchers from the University of Pennsylvania in a Nature Electronics article, A scalable ferroelectronic non-volatile memory operating at 600°C.

Ferrodiode memory devices use a 45-nanometer thin layer of a synthesized AIScN (l0.68Sc0.32N) because of its ability to retain electrical states “after an external electric field is removed,” among “other desirable properties.” Ferrodiode memory has been tested running at 600 degrees Celsius for more than 60 hours while operating at less than 15 volts.

Read more | >

MIT researchers have developed a computational approach that makes it easier to predict mutations that will lead to optimized proteins, based on a relatively small amount of data. Credit: MIT News; iStock.

MIT researchers plan to search for proteins that could be used to measure electrical activity in the brain.

To engineer proteins with useful functions, researchers usually begin with a natural protein that has a desirable function, such as emitting fluorescent light, and put it through many rounds of random mutation that eventually generate an optimized version of the protein.

Read more | >