The “computer” of 2,000 years ago fascinates us for decades. A new study points out that it might have not served at all

125 years ago, divers who were collecting sponges in the Aegean Sea, in front of the island of Anticitera, They gave with remains of an old shipwreck. Among jewels, coins and ceramic remains, there was something that caught the attention: a copper fragment of something that seemed to be a gear. It was a compendium of gears that seemed to be part of something much bigger and was baptized as’Anticitera mechanism‘. For decades it was an ignored curiosity in the archives of the National Archaeological Museum of Athens, until the researcher Derek de Solla He recovered it, studied and determined that it was an advanced Greek “computer” built at some point between 200 and 100 AC after many theories, a group of Argentine researchers have put the anti -life mechanism to see how the first computer in history behaves .. And the conclusion is that … it was nothing more than an ingenious toy. The main piece First computer For decades, this mechanism fascinated us due not only to its antiquity, but because ignorance about its function allowed the elaboration of all kinds of hypothesis. However, it was Solla who studied more in depth at the beginning, creating physical simulations about the complete state of the mechanism. Thus, and according to the researcher, the anticitera mechanism was a piece corresponding to a more complex object, composed of at least thirty bronze gears placed strategically in a wooden box with approximate dimensions of 340 x 180 x 90 millimeters. It was activated by a crank that allowed to move the gears and perform its function. Which? Well … predict astronomical positions. According to Derek’s studies, the anticitera mechanism was a planetary computer, a very cool name to describe a “computer” that allowed predicting astronomical positions, lunar phases, eclipses and, therefore, calculating cycles of the 354 -day lunar calendar or dates for sports games, among others The pieces found Through different discs, it showed the astronomical data and, supposedly, was designed to reproduce the irregular movement of the moon in its rotation thanks to specialized gears that compensated the anomalies in its trajectory. And, after that first finding, we found more and more pieces of the mechanism, which has allowed us to get an idea of ​​how it was. It was evident that it was A very advanced machine for its timebut it is also clear that it had a number of limitations. The first: however advanced Greek astronomers were, the mechanism embodied knowledge to date, so comparing it with subsequent techniques and tools is meaningless. The position of some planets is very diverted with respect to modern measurements, for example. This is logical and It does not remove merit To the device, but there were two factors that limited their precision: the mechanics itself and the manufacture of the gears. Due to wear, copper teeth could be increasingly inaccurate and, in addition, as they were manufactured by hand and not in series in a assembly chain, any deviation in the gears would affect their accuracy in the calculations. And that is what previous studies reported, such as those of Mike Edmundsbeing one of the few who has been able to directly investigate the device and lead the equipment that is responsible for your analysis. For what supposedly served Another representation of the Anticitera mechanism Testing Now, some Argentine researchers from the National University of Mar de Plata, have simulated by computer the anti -litera mechanism and their conclusion is … that it was useless. Esteban Guillermo Szigeth and Gustavo Francisco Arenas are those researchers, and have shared their conclusions in Arxiv. For their simulation, they were based on previous studies that already took into account the influence of irregular triangular teeth and the consequences of physical inaccuracies in its construction. When performing it, they realized that the triangular teeth did not seem to have a negative impact on the functioning of the mechanism, but of traffic jams in the gears when turning the crank. That, they point out, would have made it very not very practical in scientific use, practically impracticable, reducing the mechanism to an “ingenious toy.” Now, they also claim that it is the result of a simulation about what they know about the device that It was found in 1900 And that it is possible that, taking into account the necessary skill for its construction, the irregular space between the gears could be a fruit of corrosion, and not something deliberate in its design. The researchers propose that the wheels, given the slightest mismatch, would be stuck as it is extremely pointed, when we wear out or not being well balanced by the manufacturing of the time The researchers propose a series of improvements for the gears that would avoid this defect found, but also comment that it must be “cautious to assume that the measurements perfectly reflect their original values” due to what is commented: someone took too many discomfort to build that and it is unlikely that there would be something so complex, but not functional. That is why they also point out that more must be investigatedeven developing more refined techniques, to better understand the real precision and functionality of an anti -litera mechanism that still has room to fascinate us. On the fact that it was found among the remains of a wreck, the answer is much simpler: surely It was part of the loot for some Roman emperor, probably Julio César. Therefore, the anticitera mechanism remains a source of speculation and fascination due to the tremendous merit of its construction more than 2,000 years ago, but study after study, the conclusion is that the easiest thing is that it served little. Even so, researchers point out that its construction was a technological milestone. And that is something that nobody has questioned. Images | GRB16, Zde, Tony Freeth, David Higgon, Aris Dacanalis, Lindsay Macdonald, Myrto Georgakopoulou, Adam Wojcik, Mike Peel, GTS-TG In Xataka | The myth of the creative genius or why most … Read more

The first commercial biological computer in the world

The Mobile World Congress edition that is being held this week in Barcelona is full of smartphones, tablets and other devices with which users are familiar. However, too We are holding some surprises. The team that stars this article is possibly the most exotic device of this fair. It is, neither more nor less, than the first Biological computer commercially available, and has been developed by Cortical Labs, an Australian biotechnology company. Biological computing is a branch of computer science that studies, on the one hand, how we can use elements of a biological nature to process and store information. And, on the other hand, how we can inspire ourselves in the mechanisms of biological evolution to develop new algorithms that allow us to solve complex problems. If we stick to the hardware this discipline resorts to molecules derived from biological systems, such as proteins or DNA, to carry out calculations, store and recover the information. And if we enter the field of software, in particular in the artificial intelligence (AI), biological computing proposes to address some computer problems inspired by the strategy used by biology to solve some challenges. In any case, the CL1 computer, which is the biological machine that has put the team of Cortical Labs researchers ready, is framed in the biological computing branch that seeks to develop new hardware trained to process and store information. CL1 is a dream come true This sophisticated computer has been possible thanks to a large extent to the advances that have occurred in recent years in the field of Nanobiotechnology. The most precise definition of this discipline, and also the most accepted by scientists, describes it as the technology that allows us to accurately manipulate proteins to assemble more complex functional structures. The first biological computers for research had the ability to carry out calculations manipulating the RNA (ribonucleic acid) of a bacteria. In broad strokes what scientists have done so far to develop these machines was to take advantage that DNA molecules behave in the same way as a digital circuit to implement with them The same logical operations that conventional silicon processors carry out. This manipulation of DNA is possible, precisely, thanks to the advances experienced by nanobiotechnology in recent years. DNA manipulation is possible thanks to the advances experienced by nanobiotechnology in recent years Once the biological circuit was prepared, they introduced it into a coli Escherichia bacterium identical to those residing inside our stomach and intestine, and without which we could not carry out the correct digestion of food. The E. coli bacteria is simple and innocuous enough so that researchers can manipulate it effortlessly, and when manipulated DNA crosses the bacteria cell wall, the molecular machine of the cell itself translates it into a messenger RNA (MRNA). The interesting thing is that this RNM indicates to the ribosome of the cell what to do to synthesize a protein preferred by researchers. Ribosomes are the organelles or components of the cells that are responsible for the synthesis or manufacture of proteins. And now comes the most surprising: the RNM that tells the ribosome how to manufacture the protein It is only activated in the presence of a specific entryand when it does trigger the production of the protein, which is the exit. This behavior is exactly the same as a transistor. CL1, the biological computer of Cortical Labs, does not work exactly so, but the strategy in which we have just investigated is useful to understand how it works. And it does so thanks to the cultivation of real neurons in a nutrient -rich solution that allows them to grow healthy on a silicon chip that sends and receives electrical impulses. However, not everything is hardware in this machine. The person responsible for interacting directly with neurons is a biological intelligence operating system called ‘BIOS’ And it is that responsible for interacting directly with neurons is a biological intelligence operating system called ‘Bios’ (Biological Intelligence Operating System) that has been developed by Cortical Labs. This software sends information directly to neurons about its surroundings, and the latter react emitting electrical impulses. In any case, the most important thing is that CL1 neurons are programmable, and, therefore, it is possible to display code on them. However, this machine is not intended for a domestic use scenario. Its purpose is to be used by researchers to, for example, better understand how neurons process information without experimenting with animals. It can even help neuroscientific to understand how learning works in real time or the mechanisms that trigger some neurodegenerative diseases. In addition, props, the consumption of CL1 is much more restrained than that of a conventional computer. It sounds good, right? Image | Cortical Labs More information | Cortical Labs In Xataka | We already know at what speed our brain processes: just 10 bits per second

Patents linked to Switch 2 show the operation of the Joy-Con as if they were computer mice

A patent filtered and linked to switch 2originally published in August 2023, shows characteristics of the Joy-Con de Switch 2 They didn’t know each other until now. The main and most noticeable for the change that supposes with respect to the controllers of the first switch is The use of controls as if they were computer micea characteristic that Nintendo has not confirmed but suggested that it could be included among the novelties of its new console. It was in the console presentation video (the only source of official information at the time of Switch 2), where the two controls are seen sliding down a surface by the color dorsos that connect to the screen. In the patent you see a hand holding a joy-with with the SR buttons supported on the surface and eye, the thumb on the stick. It seems to be intuited that the movement sensor will be in the same back as the aforementioned buttons. They are not the only secrets that reveal these patent documents: two more traditional grips are also raised, such as PlayStation and Xbox controls, but that would include (at least one of them) sensors to be also used as mice. Japanese terms that could be translated as ‘sensor to operate as a mouse’ appear in the documents 28 times. Recall that patent documents are not guarantee of anythingonly that Nintendo has explored in that direction, but added to the clues that the presentation spot gives, the bets are more or less safe. They are small clues that round out what will be the new Nintendo presentation of the characteristics of the console, on April 6 at 3:00 p.m. Until then it is unlikely that more official confirmations of the characteristics of the Switch 2 will arrive, although elements such as this “mouse mode” are given by guaranteed. Header | Nintendo In Xataka | Switch 2 appears to be more of the same. But that is exactly the strategy that Nintendo needs right now

What is Ollama and how to use it to install artificial intelligence models on your computer

Let’s explain What is Ollama and how it works This application with which you can Install Depseek on your computeras well as other artificial intelligence models such as Llama, Phi, Mistral, Qwen, Llava or Gemma. These models will be installed locally so that you do not have to go to any website to use them, and so that everything you do stays on your computer. We are going to start the article explaining what exactly Ollama is, and everything you need to know about this program. Then, we will tell you how to install it on your computer, and then how to start using it to install an artificial intelligence model and use it. What is and how Ollama works Ollama is a program that you can install on any computer, both with Windows operating system and with macOS or GNU/Linux. It is a client of artificial intelligence models, so It is the basis on which then install an AI That you want to use. Ollama has two peculiarities. The first is that It allows you to use an AI locally. This means that instead of going to the chat page with artificial intelligence of a company, the model is installed on your computer and you use it directly without entering any website. This favors you in three ways. First because The data of everything you do stay on your PCso that no company uses them. Second because You can use AI No connection to the Internet, and third because you can skip census that has an artificial intelligence model that you are using on a website. Of course, what you can’t is to searches online. And the second particularity is that It works through your computer terminalor the system symbol in Windows. This does not have to use a separate application. When I install Ollama, then you will have to use the console of your device to install and execute in it the model you want, and the questions and the prompts you write them in the console, where you will also have the answers. How to install Ollama To start using Ollama, the first thing you have to do is Install your computer application. For that enters its website Ollama.comand click on the button Download That will appear. Now, you will go to the page where you have to Choose the operating system for which you want to lower the program. Once you have chosen it, click on the button Download. By default the web will show the system you are using, but you can download the executable of any other. When you download it, it launches the installation program. Install Ollama is very simple, you just have to click on the next button on the presentation screen, and then click on the button Install On the installation screen. How to use Ollama Once you have installed Ollama, launches the application. You will see that nothing happens, this is because you have to open your computer terminalwhich in Windows is called the system symbol. Now, before you start you have to go to the web where you will see All available AI models. The web is Ollama.com/search. Click on one of the list models To enter your card. Here you will have all the information, and you will also have a selector to choose the different versions available. When you choose a version, above the right you have the necessary command to install it and launch it in Ollama. Now, you just have to Use the command to launch the model in your terminalsuch as Ollama Run Deepseek-R1: 8b To launch the 8B version of Deepseek R1. The first time you use the command first the model will be installed, but the next ones you will already throw it directly. Remember that to do this at the terminal of your computer you must first have launched the ollama application. And after writing the command, The AI ​​model will be launched in the terminal. This is to distinguish why you see that in the terminal’s writing field now appears >>>which means that what you write will be sent to the artificial intelligence model. Now, in your computer’s command line you can Write the prompt you want Throw you to the AI ​​you have chosen, and after a few seconds the answer will begin to generate. In Xataka Basics | Prompts pages: 16 free websites and communities to find ideas for your prompts and find advice to improve them

the future where AI works for us from the computer is beginning to become a reality

OpenAI presented ChatGPT in 2022. In 2023 it arrived GPT-4 together with the innovator ‘Her’ style voice mode. In 2024 the reasoning model debuted o1. And 2025? The AI ​​company just announced Operatora product that promises to become one of the great protagonists of the year. Sam Altman and his team have presented live How does this tool work that puts people at the center of the scene? AI agents. Currently, as we’ll see, it can perform some actions in the browser, although the ambition is that, over time, it can handle more sophisticated tasks. This is what Operator looks like in action We are witnessing an initial launch. Operator is still in the research phase, but OpenAI has made a publicly available preliminary version which can be accessed from operator.chatgpt.com. The limitation? That not everyone will be able to have access to it. At the moment, it is only available for ChatGPT Pro users, the level of 200 dollars per month. The company has also noted that access to its latest product is limited to the United States. He has said that “it will take a while” to reach Europe, and that it will be available to more users “in the coming months.” The staggered launch likely responds to regulatory issues, as we have seen in the past with products that landed in the US market first. In development. Images | OpenAI In Xataka | “The objective is not to compete with ChatGPT”: we spoke with the creators of ALIA, the 100% Spanish AI, to understand its future

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.