Humans Are the Weakest Link in Signal - DTNSB 5012 ![rw-book-cover](https://wsrv.nl/?url=https%3A%2F%2Fassets.pippa.io%2Fshows%2F61954547cb03c875f7617118%2F1739388329049-4cbc8f46-1db8-425e-9fa7-e39fd5e20205.jpeg&w=100&h=100) ## Metadata - Author: [[Daily Tech News Show]] - Full Title: Humans Are the Weakest Link in Signal - DTNSB 5012 - Category: #podcasts - URL: https://share.snipd.com/episode/432886b8-763a-46b3-8673-ca7d57aeaf82 ## Document Notes ## Highlights - **Signal Leak Due to Third-party Breach** - The recent Signal leak was not due to Signal's encryption but due to a third-party archiving service, TeleMessage, being breached. - This highlights that even secure platforms can be compromised through user actions or their service configurations. Transcript: Robb Dunewood Follow up on the context of those stories, and just try to help each other understand. Today, a U.S. Government leak related to Signal. That's yet again not a weakness in Signal. And Tom explains what a CPU is. Tom Merritt I'll try. I'm Tom Merritt. And I'm Rob Dunwood. Let's start with what you need to know with the big story. Friday, Reuters reported that former U.S. National Security Advisor Mike Waltz was using an Israeli app called TeleMessage. Waltz was the official who accidentally added a journalist to a group chat on Signal. So what apps he uses are of note. And Telemessage uses modified versions of signal so people were like oh is that the version of signal that he was using telemessage is a product marketed to governments that can preserve Encrypted chats on signal whatsapp and other messaging services now you need to do this if you're in a government, because governments like the United States have laws requiring that Official communications be preserved. So archiving is not the problem. And telemessage has been around for a long time. It markets itself as a convenient and secure way to archive your government messaging. It's marketed to governments. It operates by adding itself as a third party to conversations in modified versions of the apps so that it does not break the end-to encryption, but can send messages from those chats To an archive. It's all end-to encrypted, and the archive is supposedly encrypted. Except over the weekend, a hacker who saw this Reuters story thought let's poke around and find out that hacker told 404 media that they had accessed archived chat logs from telemessages Aws instance in northern virginia now whether they weren't properly secured whether he was able to fish somehow a key or something we don't know but this hacker was able to get into the AWS instance. They could not access messaging from any individual user. They got chat logs, but the chat logs did contain parts of conversations. Now those conversations were between congressional representatives talking about cryptocurrency, but they were conversations that were supposed to be kept secret. Data related to the U.S. Customs and Border Protection Agency was in these chat logs. Coinbase information, other financial institutions information. Again, these are chat logs, so a lot of information, but not any direct messages, and none of it contained any data U.S. Cabinet members. The hackers said they wanted to see what telemessage's security was like and it took them about 15 to 20 minutes to access this archive. They did not alert telemessage before making this public because they said they were concerned it would be covered up. And once again, it's not a weakness in end-to encryption, but the actions of a user that caused this third-party service to be breached. ([Time 0:02:30](https://share.snipd.com/snip/faba81ec-2112-45b7-9806-118b0f57234e)) - **Government Archives Easily Breached** - The quick breach of TeleMessage's archived logs implies likely previous unauthorized access by advanced threat actors. - This underscores the persistent risks facing government communication archiving. Transcript: Tom Merritt Could not find reliable confirmation that TeleMessage was or was not an authorized app provider for governmental use, but it does have contracts with the US. So it could have been. So let's just assume for the sake of argument, it was. The use of tele-message wasn't the problem. A hacker over the weekend, taking 15 to 20 minutes to find the ability to get into archived logs, when you're talking about government officials, implies to me that other motivated, Advanced persistent threat type organizations probably could have done this and probably have done this as well. Robb Dunewood Yeah, I was going to say it's not a probably. This is probably. We could almost. We could almost. I don't want to say with 100% assurance because you can't be 100% about anything. But it is very, very probable that some other folks have taken a look at this just because of the speed at which it was done by somebody saying, I kind of wonder if I could get in. ([Time 0:06:34](https://share.snipd.com/snip/4e7b0dcd-93d6-405c-b850-2baf5b87e22e)) - **CPU Basics Explained** - The CPU is the central processor that executes instructions coordinating all computer operations. - Understanding CPUs clarifies that multi-core means multiple CPUs on a single chip enhancing performance. Transcript: Robb Dunewood Are the essentials for today. Let's dive a little deeper. We've been hearing a lot about chips lately. When you're hearing about people making chips or packaging chips, that can mean anything from RAM to a CPU. So Tom took some time to explain exactly what is a CPU, Tom. Tom Merritt How many times have you heard the word CPU? And how many different ways have you heard it? Probably more than you thought. Some people use CPU to refer to the main chip inside a computer. Others use it to refer to the whole motherboard. I've even heard people refer to the entire computer case ([Time 0:21:30](https://share.snipd.com/snip/656273a1-b713-4e7a-b982-453bd9084a49)) - **CPU Architectures Overview** - The von Neumann architecture separates data and instruction memory, unlike Harvard architecture which combines them. - Most modern CPUs use von Neumann design, but some embedded systems retain Harvard for specific benefits. Transcript: Tom Merritt As the CPU to differentiate it from the monitor and the keyboard. But it's actually the smallest of those. CPU stands for Central Processing Unit. And most modern computers have more than one on a single chip to improve performance. Those are called multi-core processors. When you hear the word multi-core, that means actually multiple CPUs on one chip. Fundamentally, the CPU processes instructions. Almost everything you do on a computer or a phone or a tablet or a game console is coordinated by the central processing unit. Alan Turing pioneered the idea of storing programs so that computers wouldn't have to be rewired in order to do more than one thing. Many people, including Turing himself, tried to make practical versions of this. John von Neumann published a practical outline for a stored program computer called EDVAC. That stood for Electronic Discrete Variable Automatic Computer, EDVAC. He published that on June 30th, 1945, and so he's generally credited with inventing the CPU. Several other stored program computers were finished before EDVAC, though, so he thought of it, but he didn't finish building it before the Manchester Baby and the Harvard Mark I. The Harvard Mark I represented an alternative to von Neumann's conception, though. The Harvard architecture combined memory for instructions and data together. The von Neumann architecture had separate storage and treatment of and data. ([Time 0:21:59](https://share.snipd.com/snip/1b9124f8-860b-4181-9e3e-2b15a753cb68)) - **Manchester Baby's Historic Run** - The Manchester Baby was the first stored-program computer running its first program in 1948. - It laid foundational work for modern programmable computers and CPU designs. Transcript: Tom Merritt A team at the University of Manchester, because I know you're like, wait, what was the Manchester baby? Can you talk about the baby? Team at the University of Manchester under Tom Kilburn developed a form of memory to store programs based on cathode ray tubes. Those are the same tubes that used to be used in those big televisions and monitors, you know, like you had in the 50s through the first decade of the 21st century. The Manchester baby successfully ran its first program on June 21st, 1948. It took 17 instructions and 52 minutes to produce the highest proper factor of 2 to the 18th power. So it wasn't fast for our standards, it was pretty fast for the time. The success was reported in the journal Nature, which was published in September 1948. F.C. William and T. Kilburn's electronic digital computers. He beat Alan Turing's ACE project at the National Physical Laboratory. So in 1948, Turing said, well, I'm going to go to the University of Manchester and enjoy these guys. They're fast. Kilburn oversaw the development of a series of innovative and pioneering computers known as the Manchester Computers. While there are many other computers that helped develop the form of the CPU, the Manchester computers have a direct line from the first stored program computer to the first transistor Computer. ([Time 0:23:39](https://share.snipd.com/snip/ff66dd0b-727e-42b9-b7e0-a7404399574c)) - **Transistor Revolution in CPUs** - Early transistor computers replaced vacuum tubes, enabling smaller and more efficient CPUs. - The MetroVic 950 was the first transistor computer, marking a leap in CPU evolution. Transcript: Tom Merritt Under the supervision of Kilburn, the two graduate students built and demonstrated the MetroVic 950, the world's first computer made of transistors in November 1953. It was similar to an older Manchester computer called the Mark 1, used transistors instead of using vacuum tubes. Because the prototype Metro Vic 950 didn't include a register, it was actually slower than the Mark I. But a register would be added later, making it faster, smaller, and more efficient as a CPU. Grimsdale and Webb would both get their degrees and co-write a paper on the prototype with Kilburn, which is considered a landmark paper in the evolution of the computer, and set the Template for the CPU. Of course, designs of CPUs have changed a lot over the years, but the fundamentals are pretty much the same. ([Time 0:25:52](https://share.snipd.com/snip/0f48b5c8-7261-4b39-b0f0-063d10aeee02)) - **CPU Instruction Cycle** - CPUs fetch, decode, and execute instructions via control units, registers, and ALUs. - Clock pulses time these steps, enabling operations like arithmetic and data handling. Transcript: Tom Merritt A control unit coordinates operations. A register, I just mentioned the register, the register supplies the operations and stores the results, and an arithmetic logic unit, or an ALU, performs the operations. The basic steps are fetch the instructions, which might simply be add two numbers from memory, decode those instructions, and then execute them. The control unit is just looking at the program counter, also called the pointer in x86 processors, to know what part of memory to look at for the next instruction. It fetches the instruction, then adds another program counter so it knows to go to the next instruction after the next clock cycle. That instruction is given to the binary decoder to be converted into signals used to control other parts of the CPU. The interpretation of the instruction is governed by the CPU's instruction set architecture. So when you hear someone say, well, ARM is an instruction set, that's what it means. It defines that part of the CPU. X86, which I just mentioned, is a different set of architecture, a different instruction set. Fields within each instruction determine what happens next. There's usually a field called the opcode. That indicates what op, what operation is to be performed. Then there will be other fields called operands, not to be too confusing. These tell you the values to process, you know, a 10, a 5, whatever. Could be a constant. Could be something that is sitting in the processor register. Could be from somewhere else in memory. Let's think about an instruction to add two numbers as an example. So let's say the numbers are sitting in the processor register left there after a previous instruction created them, right? This is part of a program. The last step is to execute the instructions. The control unit enables and disables parts of the CPU based on the decoded instructions. The numbers to be added flow from the register into the active parts of that arithmetic unit, the ALU, which automatically performs the desired operation. In this case, adding them together. That action is called a clock pulse. When you talk about clock cycles, that's what we're talking about. Clock speed is the number of clock pulses that a CPU can do in a short period of time. On the next clock pulse, the result of the execution, the addition of the two numbers in our example, is then outputted according to the instruction, either to the register or to be available For the next instruction, or somewhere in memory where it can be used by other programs or displayed on a screen or whatever. ([Time 0:26:39](https://share.snipd.com/snip/182d7642-cf46-4c5d-b4ef-f660cc8d3f51)) - **Modern CPU Enhancements** - Modern CPUs use techniques like caching and multithreading to improve efficiency. - Systems on a chip integrate multiple CPUs and components tightly for compact performance. Transcript: Tom Merritt The CPU just checks instructions and performs operations on numbers to put out other numbers. It's elsewhere in the computer that those numbers turn into graphics or sound or anything else. The modern CPU has some other neat tricks like caching and such to increase performance they're all varying ways of coordinating the instructions so they can be operated on the fastest Way possible for example multi-threading if you hear multi-threading it lets a cpu execute multiple threads of data at once the cpu or multiple cpus almost always exist on an integrated Circuit and that integrated circuit may also have its own memory and interface with peripherals, but the entire system of the CPU and its integrated components are called a microcontroller, Or more often these days, a system on a chip. When you hear system on a chip, it's all of those CPUs integrated together. ([Time 0:29:52](https://share.snipd.com/snip/5df3b3f6-032e-41c7-9f23-7b209d240325)) - **Choosing App Payment Options** - Developers may offer discounts for external payments to avoid Apple’s 30% cut but many will keep Apple's system for convenience. - Consumers should weigh potential savings against ease of use when choosing payment methods. Transcript: Tom Merritt Yeah, he writes, I'm listening to Friday's episode and had a quick question. This is obviously a big win for developers regarding Apple's, the judge forcing Apple to require them to allow developers to link out to alternative payment systems. This is obviously a big win for developers, Eiser writes. No longer having to pay the 30% cut to Apple for in-app payments, but from a consumer perspective, how much of a difference do you think this will make? Do you anticipate a trend where more apps start directing users to pay outside the app, or will most stick with in-app payments for convenience? Also, I'm wondering if app developers might offer better deals or promotions for payments made outside the app. I imagine Apple will continue trying to keep users inside the app ecosystem as much as possible. Curious to hear your take. Yeah, so a couple of things we know for sure. You will still have to offer the Apple in-app payment system. So I imagine a lot of people will go with that, even if there's another option, because it's easy. Companies can, under this order, say, hey, pay 30% less by clicking here and going outside the app. But again, you'll be going outside the app. That may scare a few people off, and they may be like, yeah, even though I'm going to save money, I just want to stick with what I know. And some developers won't want to maintain other payment systems. And so they'll just stick with in-app payments. I'm guessing the biggest companies, like Spotify, who already has implemented this, are going to be the big beneficiaries. Whereas smaller developers who don't want to have to develop their own payment systems may stick with the Apple payment system. Rob, what do you think? ([Time 0:33:02](https://share.snipd.com/snip/eb8e1b08-29b5-4e7f-9e52-4ddc5374a736))