By continuing to use the site or forum, you agree to the use of cookies, find out more by reading our GDPR policy

If the World Economic Forum in Davos was any indication, AI safety and security will be this year’s top priority for AI developers and enterprises alike. But first, we must overcome hype-driven distractions that siphon attention, research, and investment away from today’s most pressing AI challenges. In Davos, leaders from across the technology industry gathered, previewing innovations, and prophesying what’s to come. The excitement was impossible to ignore, and whether it is deserved or not, the annual meeting has built a reputation for exacerbating technology hype cycles and serving as an echo chamber for technology optimists. But from my perspective, there was a lot more to it. Amidst all the Davos buzz, many conversations took on the challenge of assessing critical AI challenges across development and security, and outlining a path forward. Sam Altman and Satya Nadella took on the real and present threats of LLM-generated misinformation and deep fakes -- both serious threats as nearly half of the world’s population braces for an election this year. I paneled a session alongside Yann Lecun, Max Tegmark, and Seraphina Goldfarb-Tarrant, where we discussed the need to overcome durable adoption challenges like cost and accessibility, the path to artificial general intelligence (AGI), and how we understand the utility and security of today’s AI systems. With talk of AGI and AI-powered economies continuing beyond Davos, it’s easy to lose sight of the challenges looming ahead. But to bring these long-promised AI systems and their impact to life, we first must solve the challenges of the Large Language Models (LLMs) of today and the autonomous AI systems of tomorrow. LLMs have drastically changed the makeup of enterprise technology across industries. There is no shortage of excitement. However, some have begun to feel disillusioned, questioning what AI prospects are real and which are merely hype. After all, the benefits of LLMs are matched equally by new and familiar safety and security challenges. The threat of bias and toxicity come to mind. Misinformation and security breaches threaten to disrupt elections and compromise privacy. Deep fakes are set to run rampant this year, claiming victims like Taylor Swift and President Biden with explicit content and impersonations. This is just the tip of a very large iceberg that’s yet to surface. As we forge ahead towards AGI, more challenges will be uncovered. And the solutions to today’s challenges will undoubtedly translate to future AI systems. Solutions to combat LLM-generated misinformation today might become the underpinnings of the controls used on AGI systems. Preventative measures to thwart prompt injection and data poisoning will extend far beyond LLMs, too. Putting off the questions and challenges of today ignores the reality that these AI systems are the foundations of future intelligence AI and AGI systems. Between now and an AGI future, a lot of development remains. In the quest for greater AI-driven productivity, humans remain the limiting factor. That will change in the next evolution of AI. Today’s human-to-AI systems will be phased out in favor of AI-to-AI systems as LLMs are refined and become more capable and accurate. Human-in-the-loop approaches will be replaced by light human supervision that merely ensures AI agents are operating as expected. The Internet of Agents (IoA), an interconnected system of intelligent agents with specific assignments, is the natural next step for AI. Imagine a scenario where an AI agent can detect a bug within an enterprise application’s code, assign a patch to a coding agent powered by an LLM, and push it live through an agent tasked with managing enterprise production environments. This could take several minutes. Whereas human intervention could stretch that timeline to hours or even days. Whether we like it or not, the “invisible hand” of the market will push this vision forward. As trust in AI systems builds, enterprise executives and development teams will cede control over these systems in the name of efficiency, productivity, and profitability. More in-depth details are posted on OUR FORUM.

I've written before about my nostalgia for the Windows XP- or Windows 7-era "clean install," when you could substantially improve any given pre-made PC merely by taking an official direct-from-Microsoft Windows install disk and blowing away the factory install, ridding yourself of 60-day antivirus trials, WildTangent games, outdated drivers, and whatever other software your PC maker threw on it to help subsidize its cost. You can still do that with Windows 11—in fact, it's considerably easier than it was in those '00s versions of Windows, with multiple official Microsoft-sanctioned ways to download and create an install disk, something you used to need to acquire on your own. But the resulting Windows installation is a lot less "clean" than it used to be, given the continual creep of new Microsoft apps and services into more and more parts of the core Windows experience. I frequently write about Windows, Edge, and other Microsoft-adjacent technologies as part of my day job, and I sign into my daily-use PCs with a Microsoft account, so my usage patterns may be atypical for many Ars Technica readers. But for anyone who uses Windows, Edge, or both, I thought it might be useful to detail what I'm doing to clean up a clean install of Windows, minimizing (if not eliminating) the number of annoying notifications, Microsoft services, and unasked-for apps that we have to deal with. That said, this is not a guide about creating a minimally stripped-down, telemetry-free version of Windows that removes anything other than what Microsoft allows you to remove. There are plenty of experimental hacks dedicated to that sort of thing—NTDev's Tiny11 project is one—but removing built-in Windows components can cause unexpected compatibility and security problems, and Tiny11 has historically had issues with basic table-stakes stuff like "installing security updates." The most contentious part of Windows 11's setup process relative to earlier Windows versions is that it mandates Microsoft account sign-in, with none of the readily apparent "limited account" fallbacks that existed in Windows 10. As of Windows 11 22H2, that's true of both the Home and Pro editions. There are two reasons I can think of not to sign in with a Microsoft account. The first is that you want nothing to do with a Microsoft account, thank you very much. Signing in makes you more of a target for Microsoft 365, OneDrive, or Game Pass subscription upsells since all you need to do is add them to an account that already exists, and Windows setup will offer subscriptions to each if you sign in first. I use Edge out of pragmatism rather than love—"the speed, compatibility, extensions ecosystem, and stability of Chrome with all of the Google stuff removed" is still a strong pitch, though it's gotten less so as Microsoft has pushed its own services more and more aggressively. In a vacuum, Firefox aligns better with what I want from a browser, but it just doesn't respond well to my normal tab-monster habits, despite several earnest attempts to switch. The main problem with Edge on a new install of Windows is that even more than Windows, it exists in a universe where no one would ever switch search engines or shut off any of Microsoft's value adds except by accident. Case in point: Signing in with a Microsoft account will happily sync your bookmarks, extensions, and many kinds of personal data. But settings for search engine changes, or for opting out of Microsoft services, do not sync between systems and require a fresh setup each time. Here are the Edge settings I change to maximize the browser's usefulness (and usable screen space) while minimizing annoying distractions; it involves turning off most of the stuff Microsoft has added to the Chromium version of Edge since it entered public preview five years ago. Complete information can be found on OUR FORUM.

Apple’s continued travails in China made headlines this week, with Counterpoint reporting sales down 24% inside the first six weeks of the year. But that’s not the only interesting news this week—it’s the twist behind that tale which could be a more serious issue for Apple and its iPhone in the long term, and which spells a major shift in Google’s influence over 2/3 of world’s smartphones. Despite China’s Vivo now leading the pack, toppling Apple from the top spot, the real winner is Huawei, whose sales soared 64%, putting it into second spot ahead of Apple. Even those stats ignore that Honor—the Huawei spinoff prompted by US sanctions—is broadly on par with Apple. Add Huawei and Honor together, and you would return to the kind of dominance we saw pre-Trump. This Huawei resurgence is independent of the US tech that drove its smartphone growth last time. Huawei’s initial recipe was to broadly replicate iPhone/Samsung device performance at a lower price point, and then run Android and its apps and services ecosystem to level the user experience. The US ban first removed Android and then the chipsets making all this work. Now Huawei is back with a seemingly independent supply chain and a new OS and ecosystem that is about to fully free itself from the Android world from which it was spawned. Nothing happens by accident in China. The domestic independence learning lessons from 2019-2021 is well planned. And what happens next will be just as well programmed. I warned in 2019 that “the prize for Huawei over the next decade if it can build out a successful HarmonyOS ecosystem, is huge. Not only does this deliver independence, but it also puts Huawei in control of the ‘third way’, the first major shake-up of the smartphone ecosystem in more than a decade. All of which would be bad news for Washington and California.” Five years later, and here we are. The pace of Huawei’s independent resurgence has surprised analysts. The Chinese giant has announced plans to split from Android with HarmonyOS Next. And even Nvidia has said that Huawei’s chipsets now make it a serious competitor in the AI space. The crux of my warning five years ago was as much—if not more about China—than just Huawei. The irony was that Huawei—just as TikTok has been doing since—was putting all its efforts into escaping China’s gravitational pull to be as Western as it could, to compete on a par with the US giants. The risk for the cozy smartphone world dominated by Apple’s walled garden and Google’s Android ecosystem was always that a third way, born in the world’s largest smartphone market and corralling consumers, developers, and OEMs, would shake apart the duopoly. Again—here we are. The perhaps even more interesting news this week is that Shenzhen, the city at the heart of China’s high-tech industry—including Huawei, is stepping into the fray. As reported by the South China Morning Post, Shenzhen “plans to expedite the adoption of [Huawei’s] self-developed mobile operating system HarmonyOS, heating up the platform’s rivalry with Google’s Android and Apple’s iOS in the world’s largest smartphone market.” Not only does Shenzhen plan to “boost the number of its native apps built on HarmonyOS and push for their adoption across several major sectors,” the city’s 2024 Action Plan, published last weekend, mandates that “HarmonyOS-based apps will be adopted in sectors that include government services, education, healthcare, banking and finance, transport and welfare.” Back in 2019, I suggested that “if Huawei takes a broad view, playing licensor rather that product owner, then it will pull other device manufacturers into the mix—starting with its Chinese stablemates,” and a few months later that “if Huawei can coral Chinese (and maybe non-Chinese) smartphone makers to jump from Android to its own operating system and app store, it will be a massive achievement. It will also be a serious threat to Google’s lock on the Android market.” Visit OUR FORUM for more.