Lifeboat Foundation

Could video streaming be as bad for the climate as driving a car? Calculating Internet’s hidden carbon footprint.

We are used to thinking that going digital means going green. While that is true for some activities — for example, making a video call to the other side of the ocean is better than flying there — the situation is subtler in many other cases. For example, driving a small car to the movie theatre with a friend may have lower carbon emissions than streaming the same movie alone at home.

How do we reach this conclusion? Surprisingly, making these estimates is fairly complicated.

Continue reading “Video streaming as polluting as driving? See the new calculations” »

Using a chip-based optical frequency comb, researchers transmitted almost double the global internet traffic in a single second.

Such a device could help address climate change and food scarcity, or break the Internet. Will the U.S. or China get there first?

If you’re one of the billions of people who have posted pictures of themselves on social media over the past decade, it may be time to rethink that behavior. New AI image-generation technology allows anyone to save a handful of photos (or video frames) of you, then train AI to create realistic fake photos that show you doing embarrassing or illegal things. Not everyone may be at risk, but everyone should know about it.

Photographs have always been subject to falsifications—first in darkrooms with scissors and paste and then via Adobe Photoshop through pixels. But it took a great deal of skill to pull off convincingly. Today, creating convincing photorealistic fakes has become almost trivial.

Once an AI model learns how to render someone, their image becomes a software plaything. The AI can create images of them in infinite quantities. And the AI model can be shared, allowing other people to create images of that person as well.

Silicon Valley has been obsessed with ChatGPT since it launched on Nov. 30. The clever chatbot, created by Elon Musk-founded startup OpenAI, has racked up more than a million users in its first five days and is likely to report strong engagement as people dive deeper into the charms of its impressive AI.

You can chat with it for free at chat.openai.com and ask it anything it deems appropriate. It doesn’t have access to the internet and can only respond based on the data set it was trained on, but its answers can be quite imaginative.

The Tesla and SpaceX founder is always the hero in this chatbot sensation’s stories.

Continue reading “ChatGPT Is A Huge Fan Of Elon Musk, Donald Trump And AI, But Not Google, Amazon And Apple” »

The quiet October 29 announcement of the partnership is light on details, stating that Trust Lab and In-Q-Tel — which invests in and collaborates with firms it believes will advance the mission of the CIA — will work on “a long-term project that will help identify harmful content and actors in order to safeguard the internet.” Key terms like “harmful” and “safeguard” are unexplained, but the press release goes on to say that the company will work toward “pinpointing many types of online harmful content, including toxicity and misinformation.”

Though Trust Lab’s stated mission is sympathetic and grounded in reality — online content moderation is genuinely broken — it’s difficult to imagine how aligning the startup with the CIA is compatible with Siegel’s goal of bringing greater transparency and integrity to internet governance. What would it mean, for instance, to incubate counter-misinformation technology for an agency with a vast history of perpetuating misinformation? Placing the company within the CIA’s tech pipeline also raises questions about Trust Lab’s view of who or what might be a “harmful” online, a nebulous concept that will no doubt mean something very different to the U.S. intelligence community than it means elsewhere in the internet-using world.

The ability to integrate fiber-based quantum information technology into existing optical networks would be a significant step toward applications in quantum communication. To achieve this, quantum light sources must be able to emit single photons with controllable positioning and polarization and at 1.35 and 1.55 micrometer ranges where light travels at minimum loss in existing optical fiber networks, such as telecommunications networks. This combination of features has been elusive until now, despite two decades of research efforts.

Recently, two-dimensional (2D) semiconductors have emerged as a novel platform for next-generation photonics and electronics applications. Although scientists have demonstrated 2D quantum emitters operating at the visible regime, single-photon emission in the most desirable telecom bands has never been achieved in 2D systems.

To solve this problem, researchers at Los Alamos National Laboratory developed a strain engineering protocol to deterministically create two-dimensional quantum light emitters with operating wavelength tunable across O and C telecommunication bands. The polarization of the emissions can be tuned with a magnetic field by harnessing the valley degree of freedom.

Twitter Blue plan is on halt right now but when it resumes you will have to pay an increased price of $11 per month if you subscribe from iOS, according to a report from The Information.

The report noted that the subscription plan will cost $7 per month if you purchase from the web. But it will be costlier on iOS to offset Apple’s App Store fees. Notably, Apple charges 30% fees to the developers for the first year of subscription, but it drops to 15% from the second year.

When Twitter launched its new subscription plan with a verification mark on November 9, it charged users $7.99 per month. If Twitter were to offset App Store fees, it should charge $10.38 — but the new price change of $11 sounds like a rounded-off figure.

Ablation of primary cilia in the striatum did not affect the object recognition memory, as evidenced by the more time mice spent with the novel object than the old object (Fig. 4e, f). Similarly, in the novel location recognition assay, IFT88-KO spent more time with the novel location than the old location (Fig. 4g, h), indicating a normal spatial memory. In addition, the contextual memory, measured using the fear conditioning test, was intact in the IFT88-KO mice, as revealed by the similar freezing time on the test day compared with the control mice (Fig. 4i).

The expression of the immediate-early gene cFos was used as a molecular marker of neural activity. We examined cFos immunoreactivity (number of cFos-positive cells) in structures that are parts of striatal circuits and those known to project to or receive projections from the striatum (Fig. 5a, b). First, the rostral dorsal striatum, but not the caudal striatum of IFT88-KO mice, exhibited a significant decrease of cFos immunoreactivity (Fig. 5c, d). Within the basal ganglia circuit, there was a trend for cFos immunoreactivity reductions in the output regions (SNr and the GPm), but not in the nuclei of the indirect pathway structures (lateral globus pallidus and subthalamic nucleus) (Fig. 5c, d). The main input regions to the striatum include the dopaminergic neurons of the substantia nigra pars compact (SNc) and the glutamatergic neurons of the cortices.