Ever had the experience where a particular scent reminded you of a place or flashed back a memory? Various scents and smells possess the power to take us to the past in seconds and relive the emotions associated with them. Owing to the connection of smells to emotional satisfaction, scientists are exploring the usage of Artificial Intelligence in recognizing the smells prevalent to a certain culture or organization in history. This will ensure that the historic smells are preserved and maintained for future generations, giving them the surreal experience of the past. While artificial intelligence techniques frequently brought back the structures of broken cities or revoked broken paintings in the past decade, its usage in tickling human olfactory nerves were limited. However, researchers and history scholars ensure that this crucial technology is being leveraged to its fullest potential by bringing AI into preserving historic smells.<h3 id="what-are-historic-smells">What are Historic Smells?</h3>Historic smells comprise the aromas, scents, and odours prevalent to a certain culture, social group, or civilization during its era. For example, the industrial revolution in the west made the air shift towards more of a smoke-filled incense that lurked in the air for years. The aroma of rosemary during the disappearance of the plague is linked widely to its usage in the treatment of the disease.Historic smells have garnered the attention of researchers in the past decade because the scents speak a lot about a culture?s mindset and economy. A great way to explain the connection is found in tobacco usage. The nineteenth-century witnessed the origin of this popular commodity wherein its smell was considered to be ?exotic? and luxurious among the masses. However, with its blatant usage over the years, its odour became too common, seemingly unnoticeable. By the beginning of the 21st century, the smell of tobacco started to make people in public spaces visibly uncomfortable and even nauseous. This goes on to show that scents and aromas hold a deeper meaning in history and can be used to track evolution and change the political and social environment of societies.<h3 id="how-can-ai-preserve-historic-smells"> How Can AI Preserve Historic Smells?</h3>The algorithms used in AI programs are crafted to behave like a human. Humans have a complex olfactory system helping us identify the smells and distinguish different ones, and know when to register a new one on the encounter. AI bots and programs are wired to act like a ?human nose? formed from a complex web of neural networks, hardwired to detect smells, put them in its database and recognize those smells with 100% accuracy in the future. Here is a quick process to help you understand how an AI robot or software can detect the various smells it is subjected to with perfection. Training The first step towards developing the olfactory sense in artificial intelligence bots is to train them to detect and identify odours, smells, fragrances, and all kinds of scents. This requires the algorithm to act like a ?computerized nose? that can catch smells by taking a whiff of old texts, objects and paintings, mimicking a mammalian?s olfactory organs. Feeding data The next step involves feeding the AI machines and bots the crucial information regarding the characteristics of smells so it can accurately distinguish between the various odours in an instant. This will allow the machines to produce descriptions of odours and their history precisely. Curating The team of researchers, scientists and scholars plan on working profusely to identify all kinds of smells and investigate the vast array of odours present throughout the historical references. The chemists, odour specialists, and top perfumer engineers will then evaluate this data for building and recreating different smells in compilations of an online encyclopedia. This online resource, composed of recreated scents, is then intended to be made available to all museums and heritage centres worldwide for assistance and reference. A prominent group of experts called Odeuropa is attempting to put the science behind artificial intelligence into recreation and preservation of the smells of the past. The efforts aim to store the different scents and study their significance throughout history and make the information available to all the scholars, museum lovers, and anyone fascinated by history fragments. But this is not the first or the only group to have looked deeper into the smell of histories. We are now witnessing a number of countries giving odours their due importance, like Japan who created its list of top smelling sites in the country. Olfactory exhibitions also help people take a whiff of the old scents.<h3 id="what-it-means-for-history-lovers"> What It Means For History Lovers</h3>Many countries have attempted to bring the sensory elements of their heritage and ancient culture into the picture for history enthusiasts. History is always perceived by sight and how we ?see? things. The other senses like smell, touch and taste are often neglected or paid little heed to while recreating a culture in museums or science centres. With the advent and advancements of AI-powered programs, historians can get a ?view? and experience the odours and actually ?smell? history. The ambience created by the integration of visual-olfactory senses is sure to induce more realism and authenticity in any piece of art.Moreover, having an authentic knowledge of the historic smells, the changing societal status of smells and fragrances can be studied and analyzed; all of which can aid scholars and researchers in reaching a better understanding of the reformations in social and political conditions within the heritage. The wonders of Machine learning and Artificial Intelligence are too many and prominent to ignore and pull out the ignored elements of the past. This technology will give the future generation a true sniff of the past.

Sameena Katthawala

Premium reporting on technology, startups, VC funding, and software from Techfastly tagged with AI. Read about themes, trends and strategies for AI.

Artificial Intelligence

How AI Is Helping Preserve Historic Smells?

Cybersecurity has been among the head banging issues for most organizations and rightly so with top companies losing 25$ every minute due to cybersecurity breaches. With the advances in technology, companies are putting various tools and systems to work when it comes to managing security of data and enterprise assets. However, the hackers have access to the same technology which is being used to commit countless data breaches every single day and exposing vital company details every single minute.Introduction of Artificial Intelligence is steadily changing the cybersecurity landscape and giving institutions an extra edge over the hackers and cyber sinners. Companies, irrespective of their services, are storehouses of huge number of details, records, personal data and sensitive information that must be protected at all costs. The constant bombardment of cybercrimes and data breaches have compelled the organizations worldwide to adopt and establish an automated system of detecting such threats and ensuring complete safety and security of their data. Machine learning, simulating human brain is, therefore, being looked upon as a savior that can help analysts predict the attacks beforehand and consolidate the security framework of their organization.<h2 id="artificial-intelligence-in-cybersecurity">Artificial Intelligence in Cybersecurity</h2><h3 id="threat-detection">Threat Detection</h3>The threat to organisations is huge and comes in all forms and sizes and relying on signature rules is no less than a blunder. With so many suspicious activities going on in the web, companies are bound to stumble upon threats and malicious activities that can crash complete operational functions of any company.If used properly, AI has the power to detect any unwanted messages or threats in seconds by simply scanning the web. Risk identification and management is a whole lot easier when AI based programs stand like a guard to any suspicious IP addresses, fraud websites, loopholes or similar suspicious malwares. With AI, companies can follow predictive risk management rather than reactive solutions while dealing with vulnerabilities or breaches.<h3 id="incident-prediction">Incident Prediction</h3>What can be better than dealing with threats and vulnerabilities? Detecting the future ones. Cybersecurity management is such a challenging topic that it can get overwhelming for even the experts to keep tack of all the activity that is happening over the web. Moreover, a human can get nowhere closer to a bot when it comes to scanning hundreds, in fact thousands of documents in a single day.AI operated devices can capture incidents and offer insights into the security system that not only help in dealing with the incoming threats but also expose the areas which are most likely to be affected or breached in the future. With AI, companies have a resourceful insight that can help them detect the next movement of hackers and build firewall around the data before the attack even happens.<h3 id="securing-network">Securing Network</h3>Most of the threats across mid-sized businesses are seen to arise through network pathways. The traditional ways of tracking malicious events may not prove to be as effective, especially with the cyber attackers using advanced technologies to break through network gateways.As opposed to the obsolete ways of securing network bandwidths, artificial intelligence can accelerate the process of identifying unrestricted threats from entering a network, regardless of its size. This allows huge networking endpoints to stay secured and free of any risks or potential hacks, all from a centrally regulated AI program.<h3 id="authenticated-access">Authenticated access</h3>Companies usually rely on passwords and pins to protect their internal communication channels and data networks. If any hacker is able to enter a local network where the employee?s data, files, personal information and even credit card credentials are stored, it can be a disaster for the organization.With strong AI processes like facial recognition and biometric scans, organizations can be assured of complete security at all time without having to deal with any potential security breach. AI based programs automatically identify a wrong access and instantly block it without any human intervention.<h3 id="prompt-response">Prompt Response</h3>When most organizations in the previous decade relied on getting security by taking prompt action after the threats entered their system, the present security management has opened a wide horizon. This also required a human expert to mitigate risks, perform repetitive sequences and develop quick response if ever the security framework was threatened.However, Artificial intelligence technology not only identifies the risks but also takes proper action in a matter of minutes to stop the attacks. Even as the employees doze off or are in the middle of a vacation, the threats are dealt with and diluted without any human supervision, drastically reducing the response time to unpredictable and discrete threats.<h2 id="artificial-intelligence-over-humans">Artificial Intelligence Over Humans </h2>It is quite understandable to argue that machines can never be treated as equal to humans but contrary to that belief, Artificial intelligence is by all means a front runner in the race of decision-making and learning. The threats are so huge and the data so vast that it is impossible for mere manual ways to protect even a simple network, let alone the whole organization. The signature-based approach used previously by giant companies are no longer effective in shielding the walls of digital infrastructure of a company.Firstly, artificial intelligence is able to handle and screen a huge amount of data which can never be expected of a human or even a team of hundreds. This ease of targeted screening and exposing threats from an ocean of data and information is a good enough reason for companies to make the switch to an AI based software. Secondly, AI algorithms can operate by learning over time and detect patterns over long operational cycles. This means that once AI recognizes a pattern of past attacks, it maps and clusters them and accurately detects the threat pattern in future, giving companies the assurance that the repetitive attacks or threats are not successful in causing destruction.Human analysts, thus, require the assistance of artificial intelligence technology in order to ensure integrated cybersecurity across all the touchpoints and defeating cyber threats before they cause unrepairable harm to organizations.

How is AI Changing Cybersecurity Landscape

For visually impaired children, social interaction has always been a barrier because of their inability to spot the real-time gestures made by the speakers in conversations. With experience over the years, blind children are able to develop better communication and social skills.What if there was a way to instill confidence and increase social communication among the visually impaired kids? What if there was a technology that would enable them to converse with peers in a better manner, giving them a better sense of the gestures made by the speakers?The advanced implications of AI have enabled this dream technology to be a reality as Microsoft unveiled its new device called ?PeopleLens,? which makes the lives of blind children a lot easier and socially active.<h2>What is PeopleLens' How Was It Developed?</h2>PeopleLens is an advanced AI device that allows children with low vision or vision impairments to interact and socialize among groups. The technology was developed by the leading giant Microsoft and is aimed to bring out a revolution in how the visually impaired perceive and interact with the people around them.The researchers combined studies from psychology and language and speech therapy and monitored the paralympic participants, closely observing how they were able to connect with their peers. This led to the formulation of this gadget that uses People Map and enables the children with less vision to locate and track the voices of people in their surroundings.<h2>How Does It Work?</h2>The technology works through an augmented reality system and complex machine learning algorithms that send identification and judgment signals. The people must register themselves by clicking pictures at certain angles through the phone connected to this AR device. The pictures are then converted into a series of vectors that are placed over a virtual map or mental map. This allows the child to get an idea of the position and distance of the people he/she is interacting with.The images of people sitting or interacting with the wearer are taken into account and analyzed by AI algorithms. The different cameras allow the machine to process the images and detect the pose and orientation of the speaker.The users are able to sense the position of the speaker using an actual 3D map which keeps on reconfiguring in real-time with movements and changes in the gaze of the speaker. The data or the names of the people talking are then read aloud in a spatialized audio, giving the child the exact sense of distance and the relative position of their peers.<h2>Benefits of PeopleLens</h2>As complex as the working sounds, the benefits of this new technology are immense.<ul><li>This has enabled the children with visual impairment to lead a better life by enabling them to see.</li><li>The social interaction of such children is bound to improve, and the wearers can understand their peers in a better way when they are using PeopleLens.</li><li>The children using this device also get notified of other objects like chairs or tables, apart from recognizing the people sitting in space.</li><li>The company aims to bring about a change in the way people with less vision interact with their peers, giving them a better idea of who is talking to them, who is gazing in their direction, whose heads are turned away and so on.</li><li>The teachers and parents can also communicate better and foster better relationships with this tech at hand.</li></ul><h2>Conclusion</h2>The possibilities with this thoughtful technology are endless as this technology can be modified to serve a lot many industries and sectors. This has opened up a world of opportunities for visually impaired children who struggle with their everyday lives.Microsoft has certainly taken a commendable step in this direction by opening doors for more features and allowing children with less vision to feel more connected and empowered. With the leading giant innovating in this area, we can expect to see more companies sinking their feet in these waters and putting technology to a good cause.

Premium reporting on technology, startups, VC funding, and software from Techfastly tagged with Microsoft. Read everything about Microsoft here.

Microsoft

'PeopleLens', Microsoft's Newest AI Technology, is Helping Blind People See

The gamers wait desperately as Meta announces the release of its new headset called Project Cambria in the next month. The headset is a mixed-reality device capable of performing far better than the company’s last headset, “quest 2". However, this device is not replacing Quest 2 but is likely to be a more advanced version of Meta’s best-selling headset.The company released a video of this long-awaited project back in 2021 and is planning to roll it out into the stores by October.According to the head of the parent company Mark Zuckerburg, the VR headset is built with extra features keeping in mind that the users feel an increased “social presence” while interacting with it.Project Cambria is expected to be around 800$ or more, according to the reports, which is more expensive than the former Quest 2. Speculations are ripe that the headset will be called “Meta Quest Pro,” but that is the talk for post-launch.The headset will be revealed at the Meta Annual Connect event scheduled around the next month but the gamers are already speculating and discussing the expected features and abilities of this VR device. Here is a look at some of the key features of Project Cambria that are going to be the main attractions.<h2 style="color: rgb(14, 16, 26); background: transparent; margin-top:0pt; margin-bottom:0pt;; font-weight: normal;">Key Features of Project Cambria</h2>The Project Cambria is supposedly a hybrid of AR/VR technology as it not only boasts of an amazing VR experience for the users but is also powered with enhanced cameras that connect this device to the real and virtual world equally.The main attraction of the headset is the advanced sensor technology with impressive features like face tracking and eye tracking, making it a stand-alone device that promises to give an unparalleled gaming experience to the techies. Apart from this, with the company’s efforts invested in health and fitness, Project Cambria could also feature a fitness tracking app which can be connected seamlessly with an Apple watch for measuring health.Recently, the company confirmed that the headset has special lenses called “pancake lenses”, making it a seemingly compact device but capable of generating far better 3D visuals.With specialized colour passthrough cameras on the headset, the outside surrounding area shall be depicted more realistically in real-time in order to bring the virtual and real world closer.The gamers are in for an all-new upgraded virtual avatar experience due to the presence of precise tracking of hand and eye movements by the sensors that render greater motion control. Along with detecting motion vividly, the users can give commands with just their eye movements in case of surroundings with low mobility. The eye tracking sensors used are said to be highly accurate, allowing the user to actually make eye contact with other users in virtual reality.The face tracking feature is here, so the avatars mimic the user’s facial expressions without lag, providing a surreal multiverse experience.To up the game, the headset shall include the feature of connecting to computers and preferably smartphones for better accessibility.With so many details about the sensors and trackers, the hardware specifications used inside Project Cambria is still clouded with no information.<h1 style="color: rgb(14, 16, 26); background: transparent; margin-top:0pt; margin-bottom:0pt;; font-weight: normal;">Upcoming AR Glass</h1>With VR headsets set to launch around the corner, it is fair to point out the announcement made by the CEO, Mark Zuckerberg, a few months back regarding the upcoming AR glass. This is set to hit the shelves by 2024 but the company has already built hype around the new venture.The development of Cambria is said to serve as a toolkit for developers working on the more advanced and progressive AR glasses.<h1 style="color: rgb(14, 16, 26); background: transparent; margin-top:0pt; margin-bottom:0pt;; font-weight: normal;">Conclusion</h1>Meta’s decision to reveal this long-awaited headset comes well in advance and gives the company an opportunity to test the waters before its competitor, Apple’s VR headset, hits the market next year. According to the CEO, Meta is ready to dive into the metaverse of AR and VR technology, and this announcement strengthens their efforts to create an interactive environment in the virtual world.

Tech News

Meta’s Next VR Headset to Launch in October

Premium reporting on technology, startups, VC funding, and software from Techfastly tagged with Technology Revolution. Read articles on Technological Revolution.

Technological Revolution

Parenting in New Age Digital World

For the virtual world and businesses, GPT-4 is coming with the promise of making AI-generated content more accessible and popular. 

What is GPT-4 and What Does it Mean for Businesses?

The extent of AI-powered tools and technology penetrating our day-to-day life is undoubtedly unfathomable, and just like its predecessors, GPT-4 is here to amaze the virtual world.In 2020, OpenAI released GPT-3, which leverages AI and Deep Learning to generate content. In 2021, Sam Altman, CEO of OpenAI, hinted that the company is planning to reveal GPT-4 around July-August this year. However, the company’s discreet call and no-revelation-talks have caused quite a stir among the AI fraternity, leaving hardly a trail of information on this hyped product.Even though this language model is yet to be released, there is not an ounce of doubt about its ability to excel its predecessors, and AI enthusiasts cannot stop talking about it. Here we get into the details of GPT-4, its expected features and its use cases for businesses across the globe. <h2 style="color: rgb(14, 16, 26); background: transparent; margin-top:0pt; margin-bottom:0pt;">What Exactly is GPT-4?</h2>The AI company, OpenAI has produced a series of autoregressive language models in the form of a Generative Pre-trained Transformer, more commonly referred to as GPT. These models started pouring in some time back, with the last one released in 2020 by the name GPT-3.These language models are able to generate varied kinds of texts on the internet and are responsible for popularizing the “AI-generated content”. These employ machine learning to produce auto-generated texts and queues from even a miniature input text, allowing users to generate large amounts of content.OpenAI revealed that the size of GPT-4 will almost remain the same as GPT-3, and there is enough evidence to support the fact that larger is not always better when it comes to AI-enabled products.Technically, GPT-4 is composed of 100 trillion parameters which are almost 500 times more than its forerunner, NLP. This clearly implies that the impending model is capable of performing tasks at a much faster rate and is able not just to comprehend vast amounts of data but also to generate better, quicker outcomes. <h2 style="color: rgb(14, 16, 26); background: transparent; margin-top:0pt; margin-bottom:0pt;">How is GPT-4 Useful for the Virtual Market?</h2>Just like the predictions made by the users, GPT-4 is certainly going to perform tasks in a far more powerful manner than GPT-3. This implies that the latest model is expected to be apt and prompt and will be able to understand users’ requirements better.GPT-3 worked with natural language and took some time to generate correct output, but GPT-4 shall be capable of consuming more input in the form of a higher number of symbols and generating more human-like results.For the virtual world and businesses, GPT-4 is coming with the promise of making AI-generated content more accessible and popular. This means that the internet may see an outpour of “robot-written content”. GPT-3 had already achieved auto-generated email responses and hyper-programmed text generation, and GPT-4 is trained to do it all with more agility and precision.For brands and companies leveraging this tool, content curation will become a piece of cake as long-form auto-generated content in a timely manner would yield faster results in less time and labour. The main intent of the language model is to come as close to human-like ability as possible, and with so many parameters, this upcoming product seems very true in achieving this trait.Even for the programmers and coders, code versions of GPT-4 are sure to generate codes and programs within minutes and are guaranteed to wipe out the low-skilled coders from businesses. <h2 style="color: rgb(14, 16, 26); background: transparent; margin-top:0pt; margin-bottom:0pt;">Closing Thoughts</h2>While the applications of GPT-4 are too many to predict and enlist, there is a growing concern that the popularity and accessibility of such tools are going to flood the web with auto-generated content, making it difficult to distinguish between the original and plagiarized content forms.Also, it raises another important question. Will the writer be sent home? Is AI-generated content the real king, after all? Certainly not. Even technically, GPT-4 is only a text-only model that makes it a smart content curation tool but nowhere like a human. And yes, if we are able to achieve multimodal AI(which is not nearly impossible in the upcoming decade), then the future of regular writers may be hanging on the thread. For now, these language models are here to aid the writers and make lives easier.

Write for us

Subscribe to Magazine