Group items tagged

"HTTP/2 is a more modern protocol that essentially speeds web browsing up using new ways of transporting data between the browser and server across the internet. It is backwards compatible with HTTP1.1 and uses most of the same technologies, but it is more efficient and allows servers to respond with more content than was originally requested, removing the need for the user's computer to continually send requests for more information until a website is fully loaded."

"DNS (Domain Name System) is the system which translates web addresses, such as google.com, into an IP address, such as 74.125.224.72/. A network requires the latter to actually access a website, but if the look-up system gets confused, it can give the wrong IP address - which appears to be what happened in the Iconfactory's case. Except that rather than one DNS server messing up, it was the server for the whole of China."

"Before we go deeper, it's important to note that there are two approaches to facial - or any biometric - authentication: "match on server" and "match on device." The former approach shares some of the risky aspects of facial recognition technology because it stores the details of one's most personal features-your face or your fingerprint-on a server, which is inherently insecure. There are some well-publicized examples of biometric databases being hacked, which is why so many companies are committing to only do on-device biometrics."

"Unlike other face recognition technology, their algorithm allows quick searches in big data sets. "Three million searches in a database of nearly 1bn photographs: that's hundreds of trillions of comparisons, and all on four normal servers. With this algorithm, you can search through a billion photographs in less than a second from a normal computer," said Kabakov, during an interview at the company's modest central Moscow office. The app will give you the most likely match to the face that is uploaded, as well as 10 people it thinks look similar."

"The company, which hosts 1.7m sites in the UK, said an error made during maintenance "effectively deleted" what was on some of its servers. "We can conclude that the issues faced have resulted in some data loss for some customers," the firm admitted. It started a "recovery process", but advised customers with their own data backup to rebuild their own websites."

"The malware was delivered through multiple ad networks, and used a number of vulnerabilities, including a recently-patched flaw in Microsoft's former Flash competitor Silverlight, which was discontinued in 2013. When the infected adverts hit users, they redirect the page to servers hosting the malware, which includes the widely-used (amongst cybercriminals) Angler exploit kit. That kit then attempts to find any back door it can into the target's computer, where it will install cryptolocker-style software, which encrypts the user's hard drive and demands payment in bitcoin for the keys to unlock it."

"The new feature, part of a wider set of tools covering "off-Facebook activity", will not delete anything from Facebook's servers, instead simply "disconnecting" data from an individual user's account."

"Do the potential paying customers for these large models add up to enough money to keep the servers on? That's the 13 trillion dollar question, and the answer is the difference between WorldCom and Enron, or dotcoms and cryptocurrency. Though I don't have a certain answer to this question, I am skeptical. AI decision support is potentially valuable to practitioners. Accountants might value an AI tool's ability to draft a tax return. Radiologists might value the AI's guess about whether an X-ray suggests a cancerous mass. But with AIs' tendency to "hallucinate" and confabulate, there's an increasing recognition that these AI judgments require a "human in the loop" to carefully review their judgments. In other words, an AI-supported radiologist should spend exactly the same amount of time considering your X-ray, and then see if the AI agrees with their judgment, and, if not, they should take a closer look. AI should make radiology more expensive, in order to make it more accurate. But that's not the AI business model. AI pitchmen are explicit on this score: The purpose of AI, the source of its value, is its capacity to increase productivity, which is to say, it should allow workers to do more, which will allow their bosses to fire some of them, or get each one to do more work in the same time, or both. The entire investor case for AI is "companies will buy our products so they can do more with less." It's not "business custom­ers will buy our products so their products will cost more to make, but will be of higher quality.""

"The problem with digital technology is that, for engineers, it is both intrinsically fascinating and seductively challenging, which means that they acquire a kind of tunnel vision: they are so focused on finding solutions to the technical problems that they are blinded to the wider context. At the moment, for example, the consensus-establishing processes for verifying blockchain transactions requires intensive computation, with a correspondingly heavy carbon footprint. Reducing that poses intriguing technical challenges, but focusing on them means that the engineering community isn't thinking about the governance issues raised by the technology. There may not be any central authority in a blockchain but, as Vili Lehdonvirta pointed out years ago, there are rules for what constitutes a consensus and, therefore, a question about who exactly sets those rules. The engineers? The owners of the biggest supercomputers on the chain? Goldman Sachs? These are ultimately political questions, not technical ones."

"But when you dive into the code of Meitu, that's where things get interesting. Security researchers have jumped in to assess the photo editing app and found that it was indeed collecting information, including a phone's IMEI number (a handset's unique ID number), and sending it back to remote servers:"

""The attraction for jihadists of these platforms is that the developers of these decentralised platforms have no way of acting against content that is stored on user-operated servers or content that's shared across a dispersed network of users, " BBC Monitoring senior jihadi specialist Mina Al-Lami said. "It's really all about privacy, freedom and encryption."

"The flaw, dubbed "Log4Shell", may be the worst computer vulnerability discovered in years. It was uncovered in an open-source logging tool that is ubiquitous in cloud servers and enterprise software used across the industry and the government. Unless it is fixed, it grants criminals, spies and programming novices alike, easy access to internal networks where they can loot valuable data, plant malware, erase crucial information and much more."

The internet's largest blogging platform went offline for nearly two hours silencing more than 10 million blogs.

"Trying to shake humanity from its complacency about this, Yudkowsky published an op-ed in Time last spring that advised shutting down the computer farms where AIs are grown and trained. In clear, crisp prose, he speculated about the possible need for airstrikes targeted on datacentres; perhaps even nuclear exchange. Was he on to something?"