The Comunicano for Xmas Day, December 25th 2024
Ho Ho Ho
Hey, everybody! Happy Holidays! Merry Christmas, Happy Hanukkah, Festive Kwanzaa, whatever you're celebrating—let's just agree to celebrate surviving 2024, huh? I mean, it’s been a year. Let’s give ourselves a round of applause for making it through—without a single robot overthrow, by the way. Let’s keep that streak going!
Now, before we dive in, you might notice something a little different about today’s newsletter. We’re trying a slightly new look, a new format. You know, like when Coca-Cola tried “New Coke”—but this time, we’re hoping it works out better!
We’ve got some great updates for you from The Comunicano. Think of it as your tech-savvy friend who reads everything so you don’t have to.
First up, AI is taking exams tougher than your in-laws’ grilling at Christmas dinner. There’s this thing called “Humanity’s Last Exam.” I know, sounds dramatic, right? Makes me wonder—if the robots are that smart, maybe they can help me finally assemble that IKEA dresser without 47 leftover screws. Anybody else feel like they’re building Swedish modern art?
And have you heard about synthetic data? It’s data...that’s not real! Like the salad I tell my doctor I eat. But seriously, it’s going to change AI training and maybe even help your grandma finally figure out her smartphone. Speaking of which, smartphones in 2025? They’re all looking the same. I guess if Apple and Samsung have their way, the only innovation will be how much you overpay.
Oh, and for the grand finale, Boston Dynamics’ robot can do a backflip now. Impressive! But here’s my question: can it put the star on top of a Christmas tree without knocking it over? Let’s work on the basics, folks!
That’s our new Comunicano-inspired vibe. Thanks for tuning in—and remember, it’s not the tech that makes the holidays special, it’s the people. Well, the people and the cookies.
Happy Holidays, from THE COMUNICANO!!!
Andy Abramson
AI Watch
AI Evaluations and Safety Concerns (Time.com)—As AI systems rapidly advance, traditional evaluations like the SATs and bar exams are becoming inadequate measures of their capabilities. In response, organizations are developing more challenging assessments, such as Epoch AI's FrontierMath and the forthcoming "Humanity's Last Exam," to better gauge AI performance. However, creating scientifically rigorous and practical tests remains complex, with concerns about data contamination and the potential for AI models to "game" these evaluations. Despite these challenges, enhanced testing is essential to understand and mitigate risks associated with advanced AI, including cybersecurity threats and bioterrorism. These evaluations are key to guiding policy and regulations.
Read more here
Synthetic Data: Promise and Perils (TechCrunch)—Synthetic data—artificially generated datasets—offers a revolutionary approach to training AI models while addressing privacy concerns and biases. This method is gaining traction in fields like healthcare and finance, where real-world data is scarce or sensitive. However, challenges like replicating real-world variability and avoiding unintended consequences persist. Advocates argue that synthetic data can democratize AI access, but critics warn of overreliance and potential unintended outcomes. This duality highlights the need for careful implementation as synthetic data reshapes the AI training landscape.
AI Use Cases Expected to Explode in 2025 (Fast Company)—Artificial intelligence is set to transform industries at an even larger scale in 2025, with emerging use cases redefining healthcare, logistics, customer service, and creative industries. AI-powered systems are anticipated to manage more complex operations, such as real-time supply chain adjustments and personalized healthcare diagnostics. As businesses integrate generative AI and predictive analytics, the technology's role will shift from efficiency tools to innovation drivers. However, scaling these solutions demands robust data governance, ethical frameworks, and workforce upskilling. The year 2025 could mark a pivotal moment in AI's integration into everyday business operations.
Apple Watch
Apple Defends Revenue from Google Payments (Reuters)—Apple is seeking to intervene in Google's upcoming U.S. antitrust trial concerning online search to protect its lucrative revenue-sharing agreements, which designate Google as the default search engine on Apple's Safari browser. These agreements reportedly earned Apple about $20 billion in 2022. Apple argues it cannot rely on Google to defend its interests, especially as Google faces potential business unit separations. The trial, scheduled for April, will see prosecutors advocating for measures to enhance competition, possibly including divesting Google's Chrome browser and Android OS. Apple plans to call witnesses to support its case, highlighting the trial's potential to significantly alter the online search landscape.
Read more here
Google Watch
Google to Delete Inactive Accounts (Forbes)—Google has announced plans to delete inactive Google and Gmail accounts that haven't been used for two years, starting from December 2024. This move aims to enhance security by removing potentially vulnerable accounts. Users are advised to log in to their accounts to prevent deletion, especially if they have important data stored. Google has been sending multiple warning emails to affected users, urging them to take action to preserve their accounts and data.
Read more here
Smartphone Watch
North American Smartphone Market in 2025 (Android Central)—The North American smartphone market is facing challenges as it heads into 2025. With fewer players, potential tariff impacts, and design homogenization, smartphones risk becoming mere commodities. The dominance of major brands like Apple, Samsung, Google, and Motorola, coupled with carrier-driven sales, limits consumer choice. Design convergence towards iPhone-like aesthetics and software features further diminishes differentiation. These trends suggest that 2025's smartphones may lack the innovation and diversity that once characterized the market, leading to a more monotonous consumer experience.
Read more here
Robot Watch
Boston Dynamics' Atlas Performs Backflip (TechCrunch)—Boston Dynamics has showcased its electric-powered Atlas robot performing a stunning backflip in a new video. The demonstration highlights Atlas's advanced agility and balance, reinforcing the progress in humanoid robotics. This latest achievement underscores Boston Dynamics' commitment to pushing the boundaries of robotic capabilities, with implications for future applications in labor, disaster response, and more. While the backflip is primarily a display of technical prowess, it represents a significant step in the evolution of dynamic, lifelike robots.
Read more here
OpenAI Explores Humanoid Robots Integration (Gadgets360)
OpenAI is reportedly in discussions to develop humanoid robots integrated with its advanced AI models, aiming to bridge artificial intelligence with physical embodiment. The initiative envisions robots capable of performing complex tasks across industries, from manufacturing to elder care, leveraging OpenAI's expertise in machine learning and robotics. This ambitious move reflects the growing push to create AI-powered robots that adapt dynamically to human environments, raising both opportunities for innovation and ethical considerations about AI-human interaction. While details remain sparse, this signals OpenAI's intent to expand beyond software and into hardware-driven AI applications.
The People Building the Future of Robots (MIT Technology Review)—Robots may dazzle with their capabilities, but the unsung heroes are the human engineers, data annotators, and testers shaping their intelligence. Behind every robot lies hours of meticulous programming and decision-making to align machine learning systems with real-world needs. This piece explores the lives of individuals working on integrating AI into robotics, from crafting precise algorithms to ensuring ethical AI deployment. Their work underscores the intricate interplay between human creativity and computational precision, offering a glimpse into how robotics teams navigate the challenges of making machines smarter and more empathetic.
Data Center Watch
Navajo Land Powers Arizona Data Centers (Washington Post)—Arizona’s sprawling data centers are increasingly relying on renewable energy from Navajo land, with partnerships between tribal energy initiatives and utility giants like APS and SRP. Navajo Power, a key player, is driving solar energy projects to meet the growing demand for clean power from tech companies while promoting economic development for the Navajo Nation. These collaborations highlight the delicate balance of providing sustainable solutions for energy-intensive data centers and empowering Indigenous communities through clean energy innovation, though challenges persist in scaling these efforts equitably.
Security Watch
FCC's Rip-and-Replace Mandate Dominates NDAA (The Record)—The National Defense Authorization Act (NDAA) includes stringent measures requiring the removal of telecommunications equipment from Chinese companies like Huawei and ZTE, citing national security concerns. The "rip-and-replace" initiative demands that U.S. telecom networks overhaul their infrastructure to mitigate risks tied to foreign surveillance and cyberattacks. While this policy reinforces the government's stance on securing domestic communications, it also presents logistical and financial hurdles for smaller providers reliant on federal funding to comply. The NDAA solidifies America's push to safeguard its digital landscape against geopolitical threats.
EV Watch
Mercedes-Benz Advances with Level 3 Autonomy (HyperBeast)—Mercedes-Benz has secured regulatory approval for its Level 3 autonomous driving system, positioning itself as a leader in semi-autonomous vehicle technology. This achievement enables drivers to relinquish control in certain conditions, such as highway traffic, while the car handles navigation and safety. Unlike fully autonomous vehicles, Level 3 systems still require driver oversight but represent a significant leap toward automotive autonomy. The development signals a new era of driving convenience, sparking competition and raising questions about regulatory frameworks and user readiness.
