Here is the daily technology #threadcast for 5/9/25. The goal is to make this a technology "reddit".
Drop all question, comments, and articles relating to #technology and the future. The goal is make it a technology center.
Here is the daily technology #threadcast for 5/9/25. The goal is to make this a technology "reddit".
Drop all question, comments, and articles relating to #technology and the future. The goal is make it a technology center.
!summarize #ai #jobs #generativeai
Paragraphs
In another positive sign for the company, its first-quarter capacity utilization— the percentage of total available manufacturing capacity that is being used at any given time— reached 89.6%, up 4.1% quarter on quarter.
"SMIC's nearly 90% utilization rate reflects strong domestic demand for semiconductors, likely driven by smartphone and consumer electronics production," said Ray Wang, a Washington-based semiconductor and technology analyst, adding that the demand was also reflected in the company's strong quarterly revenue growth.
Meanwhile, the company said in the earnings call that it is "currently in an important period of capacity construction, roll out, and continuously increasing market share."
However, SMIC's first-quarter research and development spending decreased to $148.9 million, down from $217 million in the previous quarter.
In the earnings call, the company also said it would closely monitor the potential impacts of the U.S.-China trade war on its demand, noting a lack of visibility for the second half of the year.
Phelix Lee, an equity analyst for Morningstar focused on semiconductors, told CNBC that the impacts of U.S. tariffs on SMIC are limited due to most of its revenue coming from Chinese customers.
While U.S. customers make up about 8-15% of revenue on a quarterly basis, the chips usually remain and are consumed in Chinese products and end users, he said.
"There could be some disruption to chemical, gas, and equipment supply; but the firm is working on alternatives in China and other non-U.S. regions," he added.
For the second quarter, Lyft said it anticipates rides growth in the mid-teens from a year ago. Gross bookings are expected to range between $4.41 billion to $4.57 billion. Analysts polled by FactSet had forecast gross booking of $4.48 billion.
Lyft reported $280.7 million in free cash flows for the first quarter, which topped a $136.3 million estimate from StreetAccount.
Last month, Lyft announced an expansion into Europe with the nearly $200 million acquisition of Germany-based taxi app FreeNow.
Pinterest's finance chief Julia Brau Donnelly told analysts during an earnings call that while the company is "not immune to the macro environment," executives are optimistic about its advertising products and that its overall business is healthy.
Still, Donnelly said "there have been small pockets of spend that have been impacted by tariffs in recent weeks," pointing to unspecified Asia-based e-commerce retailers curbing their digital advertising spend in light of the recent ending of the de minimis trade loophole.
Bill Ready, Pinterest's CEO, told analysts that executives have focused on morphing the company "from a platform with declining users and modest revenue growth a few years ago into a secular share taker with a more resilient business than ever before."
In his YouTube video titled “How I Built a Real Movie Star Robot,” DeWitt shows the entire process. He 3D printed the parts, carefully hand-sanded each piece to smooth perfection, and applied a primer coat. Then came the hand-painting. But he didn’t stop there.
To give Newt a realistic, lived-in look, DeWitt used special weathering techniques. This gave the robot a slightly worn and rugged appearance—making it feel like part of the real world, not just a fresh-off-the-shelf prop.
Limitless nuclear power
GLE anticipates producing hundreds of kilograms of enriched uranium during the TRL-6 testing period.
The output and system behavior will directly inform deployment pathways for the PLEF. It will support pre-licensing submissions to the Nuclear Regulatory Commission, including the formal safety case expected this summer.
Laser enrichment, as implemented through the SILEX process, offers substantial gains in enrichment efficiency by leveraging highly selective laser excitation of uranium hexafluoride isotopes.
The process enables the isolation of U-235 isotopes with greater energy efficiency and process control than traditional methods such as gaseous diffusion and centrifuge technology.
Given the increasing diversification of reactor fuel needs, especially with the emergence of small modular reactors (SMRs) and advanced reactor platforms, the demand for flexible, high-assay enrichment capabilities is projected to expand.
Elaborating on the technological underpinnings, Lenovo CTO Tolga Kurtoglu detailed the robot’s intelligence as being driven by a sophisticated three-layered architecture.
This framework facilitates proactive device interaction, seamless data integration across Lenovo’s ecosystem, and advanced autonomous task execution, all while adhering to stringent security and privacy protocols.
Powering this new class of intelligent agent is Lenovo’s enterprise intelligence strategy, which leverages a four-layer “device-edge-cloud-network” architecture. This hybrid computing foundation integrates both public and private clouds and facilitates efficient data flow from collection and storage to processing and AI model training via intelligent scheduling.
Shares in Chinese chipmaker SMIC drop nearly 7% after earnings miss
Shares of Semiconductor Manufacturing International Corporation, China's largest contract chip maker, fell nearly 7% after an earnings miss.
Shares of Semiconductor Manufacturing International Corporation, China's largest contract chip maker, fell nearly 7% Friday after its first-quarter earnings missed estimates.
After trading on Thursday, the company reported a first-quarter revenue of $2.24 billion, up about 28% from a year earlier. Meanwhile, profit attributable to shareholders surged 162% year on year to $188 million.
However, both figures missed LSEG mean estimates of $2.34 billion in revenue and $225.1 million in net income, as well as the company's own forecasts.
During an earnings call Friday, an SMIC representative said the earnings missed original guidance due to "production fluctuations" which sent blended average selling prices falling. This impact is expected to extend into the second quarter, they added.
For the current quarter, the chipmaker forecasted revenue to fall 4% to 6% sequentially. Gross margin is also expected to fall within the range of 18% to 20%, compared to 22.5% in the first quarter.
Still, the first quarter saw SMIC's wafer shipments increase by 15% from the previous quarter and by about 28% year-on-year.
In the earnings call, SMIC attributed that growth to customer shipment pull in, brought by changes in geopolitics and increased demand driven by government policies such as domestic trade-in programs and consumption subsidies.
Celsius CEO Alex Mashinsky sentenced to 12 years in multi-billion-dollar crypto fraud case
Mashinsky's downfall mirrors the fate of FTX founder Sam Bankman-Fried, Binance's Changpeng Zhao and Do Kwon of Terraform Labs.
Alexander Mashinsky, the former CEO of Celsius Network, was sentenced to 12 years in prison on Thursday after pleading guilty to two counts of fraud, a dramatic fall for the leader of a company once hailed as the "bank" of the crypto industry.
Standing before U.S. District Judge John G. Koeltl in Manhattan's Southern District, Mashinsky faced the consequences of what prosecutors described as a sweeping scheme to defraud investors.
In December he pleaded guilty to commodities fraud and a scheme to manipulate the Celsius token.
His sentencing took place in courtroom 14A at 500 Pearl Street — a venue that has seen several crypto executives-turned-felons.
Mashinsky's legal troubles began in 2023 when he was arrested on charges of securities, commodities, and wire fraud, just as Celsius reached a $4.7 billion settlement with the Federal Trade Commission — one of the largest in the FTC's history.
The settlement, which remains contingent on Celsius returning what remains of customer assets in bankruptcy proceedings, underscored the magnitude of the fraud.
Prosecutors accused Mashinsky of misleading investors about the safety and profitability of Celsius's yield-generating platform while secretly selling off tens of millions of dollars in personal holdings.
!summarize #transformers #learning #ai
YouTuber builds Wall-E-like cute robot sidekick from scratch for sci-fi film
From 3D-printed parts to Arduino-powered insides, Newt was built to perform real-time actions on camera.
The first time Davis DeWitt saw the character concept for Newt, he may have thought of Wall-E. But unlike Pixar’s animated robot, Newt is real—and ready for the big screen. This functional robot was designed and built entirely from scratch by DeWitt, a YouTuber and the founder of Backhaul Studios.
Newt will star in the upcoming sci-fi adventure The Lightning Code, which is scheduled to hit theaters on May 9, 2025. The family-friendly movie follows a teenage inventor who discovers a way to create limitless energy. But after corporate villains steal his invention, the teen must join forces with friends—and a charming little robot—to stop them from turning his work into a weapon.
Director Kali Bailey had one rule from the start: no full CGI for the robot. She wanted the actors to engage with a real character on set. So she reached out to Backhaul Studios, giving DeWitt a once-in-a-lifetime opportunity—building a robot that would play a starring role in a Hollywood movie.
Bringing a robot character to life
To meet the needs of the movie’s storyline, Newt couldn’t just look good. The robot had to act, move, and feel like a living character. That meant the build started with a lot of creative brainstorming. The team worked through concept sketches to imagine what Newt would look like in the film. From there, DeWitt began creating digital 3D CAD models to finalize the design.
After several rounds of revisions, they settled on a look—Newt would have a unique projector-style head. With the digital design locked in, it was time to bring the robot into the real world.
Endless nuclear power: US tests world’s most powerful laser to enrich uranium
The TRL-6 phase constitutes a high-fidelity operational trial under conditions representative of scaled commercial deployment.
Global Laser Enrichment (GLE), holder of exclusive commercial rights to the SILEX uranium enrichment process, has commenced Technology Readiness Level 6 (TRL-6) demonstration testing at its Test Loop facility in Wilmington, North Carolina.
This marks an important milestone in the operational maturation of third-generation enrichment capabilities and represents a strategic inflection point for US-based nuclear fuel resilience.
The TRL-6 phase constitutes a high-fidelity operational trial under conditions representative of scaled commercial deployment.
Laser uranium enrichment
According to GLE Chief Executive Officer Stephen Long, this milestone follows extensive system integration, engineering refinement, and iterative operational testing.
“This demonstration campaign is the culmination of years of technical advancement and operational discipline. We are now entering a phase of focused optimization, with iterative system tuning and data accumulation to fulfill the validation requirements of TRL-6,” said Long.
Key modifications have been integrated into the Test Loop to increase performance reliability.
These enhancements validate scale enrichment efficiency and inform downstream manufacturing processes for Kentucky’s full-scale Paducah Laser Enrichment Facility (PLEF).
Independent engineering assessment—contracted by GLE’s joint venture stakeholders, Silex Systems Ltd, and Cameco Corporation, which will monitor and verify all test outcomes.
China’s new humanoid robot performs Tai Chi, could be Lenovo’s ‘silicon employee’
With the unveiling of LeXiang No.1, Lenovo signals its intent to move beyond its established hardware-centric model.
Global technology leader Lenovo has signaled a major strategic shift by unveiling its first humanoid robot, LeXiang No.1, at the company’s Tech World 2025 conference in Shanghai.
Positioning the robot as a future “silicon employee,” Lenovo’s move underscores its accelerating push into the fields of embodied intelligence and artificial intelligence.
The introduction of LeXiang No.1 highlights Lenovo’s ambition to evolve beyond its traditional hardware focus. Speaking at the Wednesday event, Lenovo CEO Yang Yuanqing articulated this vision.
“Today, many still see Lenovo as a device company, but in the future, intelligent agents may become the core of our business,” he said, as reported by Technode, while emphasizing that AI is providing unprecedented clarity for the company’s path forward, a theme central to Tech World 2025, which is focused on “AI for Innovation and Productivity.”
Core technological architecture
Attendees were given a glimpse into the robot’s capabilities through a live demonstration. LeXiang No.1 performed a tai chi routine, showcasing its fluid movements and balance control.
In a more practical display, it accessed Lenovo’s comprehensive business data to answer questions during a Q&A session, hinting at its potential for sophisticated enterprise tasks.
Apple Takes a Swing at Google in Mobile Safari
Apple's gearing up to launch AI-powered search in mobile Safari later this year, taking direct aim at Google’s search dominance. SVP Eddy Cue says the AI will help users “do more than just find stuff”—think smarter suggestions and deeper answers. It’s part of Apple’s broader push into AI tools expected at WWDC in June. With iPhones in over a billion pockets, even a small shift could shake up the search game big time.
#apple #ai #safari #search #technology
> S👁️URCE <
Meet Vulcan: Amazon's Robot With a Human Touch
Amazon just rolled out Vulcan, a warehouse robot with a sense of touch—yes, really. It can gently grab everything from earbuds to shampoo, handling 75% of the million-plus items at the Spokane, WA site. Think of it like a super tireless co-worker: it lifts up to 8 pounds, runs 20 hours a day, but isn’t here to take jobs—Amazon says it’s paving the way for higher-skilled roles instead.
#automation #robots #futureofwork #warehouse #technology
> S👁️URCE <
Google Bets Big on Next-Gen Nuclear
Google just teamed up with Elementl Power to develop three nuclear reactor sites, each packing enough juice to power a small city—600 megawatts or more. The deal gives Google the right to buy the clean energy once the plants are live. Elementl hasn’t picked the exact reactor tech yet—they’re waiting to go with the most ready-to-build option when it’s time to break ground.
#google #nuclearenergy #cleanpower #futuretech #technology
> S👁️URCE <
Figma’s New AI Tool Wants to Read Your Mind
Figma just unveiled Make, an AI-powered design tool that turns your ideas into websites and apps—kind of like having a super-talented coder in your pocket. It’s built on Claude 3.7 Sonnet from Anthropic and aims to simplify complex design work into a “vibe-coding” experience. Pricing starts at $16/month per user (annual only), and yep, no free version. Still in testing, and no release date yet, but this could reshape how creatives build fast.
#figma #ai #designtools #startups #technology
> S👁️URCE <
The Hidden Costs of Working with AI
Building with LLMs isn't like coding apps—it’s more like training a wild animal. You don’t write logic, you guide probabilities. That means testing, failing, tweaking… and doing it all over again. You’ll need humans in the loop, smart evaluation tools, and a mindset ready for surprises. And unlike traditional software, LLMs rack up ongoing costs—even after launch. Bottom line? Future-proofing AI is a myth. You learn by doing. A lot.
#ai #llm #machinelearning #futureofwork #technology
> S👁️URCE <
Maxing Out Your GPU Like a Pro
If you've ever felt like your pricey GPU is just twiddling its digital thumbs, you're not alone. This guide dives into how to squeeze every drop of performance from it—especially when running neural networks. Think of it like making sure every seat on a speeding train is filled. In just 20 minutes, you'll learn how to track, tune, and truly use the hardware you paid for.
#machinelearning #deeplearning #gpuoptimization #neuralnetworks #technology
> S👁️URCE <
Netflix Just Went Full TikTok on Mobile
Netflix is testing a TikTok-style feed in its mobile app, letting you scroll short clips from shows and movies. If something grabs you, you can dive in instantly, save it for later, or share it with a friend. Think of it like channel surfing in 2025—only vertical, fast, and on your phone. It’s Netflix’s latest move to keep us hooked in a world where attention spans are shorter than a coffee break.
#streaming #netflix #mobileapps #innovation #technology
> S👁️URCE <
Claude Just Got a Superpower: Web Search via API
Claude can now browse the internet in real time, thanks to Anthropic’s new API update. That means apps using Claude can finally pull in live info—think weather updates, breaking news, or stock prices without lag. It's like giving your AI assistant a direct line to the internet, instead of relying on its memory. This shift makes AI tools way more useful for stuff that changes by the minute.
#ai #anthropic #websearch #realtimedata #technology
> S👁️URCE <
AI Battles Human in Street View Showdown
Can an AI out-navigate a geography wizard? Researcher Sampatt put O3 to the test in a head-to-head GeoGuessr duel against a seasoned pro. O3 flexed some serious map smarts—spotting clues in signs, terrain, and architecture—but still couldn’t outguess the human consistently. It’s a glimpse into how far AI has come in reading the world like we do—and how far it still has to go when things get messy and real.
#geoguessr #aiagents #mapgames #techshowdown #technology
> S👁️URCE <
How PyTorch Went From Lab Tool to AI Powerhouse
Once just a research tool, PyTorch now fuels some of the most advanced AI models—think of it as the engine under the hood of ChatGPT-like systems. In 2022, the PyTorch Foundation grew to include tools that make building and scaling AI smoother, like adding gears to a fast car. It's now the go-to platform for developers chasing the next big AI breakthrough.
#pytorch #deeplearning #opensource #genai #technology
> S👁️URCE <
Meta's New AI App Lets You Share Your AI Moments
Meta just launched an AI app with a fun twist: the Discover feed, where you can share your AI interactions, like, comment, and remix posts! It’s built on the Llama 4 model and offers voice interactions in select regions. This replaces the View app for Meta's Ray-Ban glasses. It's Meta's way of making AI more social and less mysterious, so it’s not just a tech tool—it’s something you can share and connect over.
#meta #AI #socialmedia #technology
> S👁️URCE <
AI Shopping Assistants Are Here to Make Your Life Easier!
Mastercard's new AI program is revolutionizing online shopping by handling the hunt for products. No more wasting time browsing – AI agents find exactly what you need, faster. But don’t worry, you’re still the boss. You get the final say on purchases. Think of it like having a super-efficient assistant, without giving up control. It’s a smart upgrade for your shopping routine.
#mastercard #ecommerce #shoppingassistant #technology
> S👁️URCE <
Meta Launches AI Defenders Program to Protect LLaMA Models
Meta just kicked off its AI Defenders Program, a new initiative aimed at preventing the misuse of its LLaMA models. This program helps developers spot harmful AI use and ensure the models are being used responsibly. Think of it as a digital security system for AI—ensuring that what’s built to help doesn’t get misused in the wrong hands. Big move for protecting AI tools!
#meta #AIdefenders #LLaMA #technology
> S👁️URCE <
Meta's Smart Glasses May Soon Recognize Faces
Meta is developing a 'super sensing' facial recognition tech for its smart glasses. This feature will identify people by name and track users' activities, but it’s opt-in only. The company is also working on two new pairs of glasses while revising privacy policies and assessing safety risks. Could these glasses be the next big thing in wearable tech? Only time will tell.
#Meta #SmartGlasses #Privacy #Technology
> S👁️URCE <
AURORA EXPANDS DRIVERLESS TRUCKING ROUTES & NIGHT DRIVING
Aurora Innovation is ramping up its driverless trucking game. Starting in late 2025, the company will expand its routes from Dallas to Houston, El Paso, and Phoenix. They’ll also launch nighttime and adverse weather driving, doubling their operational hours. Currently, they run self-driving trucks with safety operators, but this next phase is a major step toward full autonomy in challenging conditions.
#aurorainnovation #driverlesstrucking #autonomousvehicles #technology
> S👁️URCE <
Saola’s Genome Sequenced—Could Save This 'Asian Unicorn'
The saola, a rare bovine found in Vietnam and Laos, has been virtually invisible to science since its discovery in 1993. With only 50-300 left in the wild, it's one of the world’s most endangered species. But now, an international team has sequenced its genome, uncovering genetic diversity that could kickstart a breeding program. This breakthrough could be a game-changer for the species’ survival. Let's hope it leads to a comeback for this "Asian unicorn."
#saola #conservation #genomics #endangeredspecies #technology
> S👁️URCE <
Finding High-Information Density UIs: A Hacker News Dive
If you're hunting for high-information density UIs, a Hacker News thread just dropped a bunch of gems. Pro creatives often rely on certain programs daily—these tools are prime examples. Also, command line system monitoring tools show how dense data can be presented without overwhelming. It's all about making data both accessible and powerful. Dive into these resources to learn from the pros!
#uidesign #userinterface #productivity #technology
> S👁️URCE <
AI in Cybercrime: A Game Changer
AI isn't just a tool for cybersecurity; it's also lowering the bar for cybercriminals. At the RSA Conference, experts showed how AI helps automate attacks and even offers “AI-as-a-Service” to bad actors. As cybercrime evolves, we’ll need smarter, AI-driven defenses and global teamwork to keep up. Imagine the stakes—AI's power is in the hands of both the good and the bad.
#cybersecurity #ai #rsa2025 #cybercrime #technology
> S👁️URCE <
THE LEADERBOARD ILLUSION: WHY CHATBOTS AREN'T GETTING THE FULL PICTURE
Chatbot Arena’s benchmarking isn’t as neutral as it seems. With private tests and selective data access, major players like Google and OpenAI have the upper hand, leaving open-source models behind. This bias skews results, leading to overfitting—where models perform well on familiar tests but fail in real-world use. It’s a reminder that data access can make or break a model’s progress. #chatbots #AI #openAI #Google #technology
> S👁️URCE <
OSMOSIS: AI SELF-IMPROVEMENT MADE EASY
Osmosis is revolutionizing self-improvement with real-time reinforcement learning. Their lightweight, open-sourced model matches the performance of top-tier models for MCP, and it’s simple enough to run locally. Whether you’re into tech or personal growth, this tool empowers you to boost your own skills and knowledge, making AI-driven learning accessible to everyone. Talk about leveling up!
#osmois #reinforcementlearning #opensource #technology
> S👁️URCE <
AI-GENERATED CODE: A SUPPLY CHAIN NIGHTMARE IN THE MAKING
AI-generated code might sound like a tech dream, but it could lead to a nightmare for software security. A recent study revealed 19.7% of dependencies from AI-generated code were completely fabricated. This exposes systems to supply-chain attacks like dependency confusion. Open-source models hallucinate more often than commercial ones, with JavaScript errors more common than Python. This is a serious wake-up call for developers relying on AI to code.
#security #ai #softwaredevelopment #technology
> S👁️URCE <
Hugging Face Unveils Cloud-Based AI Agent
Hugging Face just dropped the Open Computer Agent, a cloud-hosted AI tool that can handle simple tasks. While it still struggles with complex ones (like flight bookings), it shows how open AI models are changing workflows. A recent KPMG survey revealed that 65% of companies are testing out AI agents, and the market is set for massive growth. This tool is a glimpse into a future where AI assistants do more, faster.
#ai #machinelearning #innovation #business #technology
> S👁️URCE <