Midjourney AI predicts what the ‘last selfie ever taken’ will look like. The backdrop is a barren landscape with dark smoke and untamed fire in the background. The last selfie is hollow eyes and puckered skin.
Midjourney AI generator allows users to imagine alternative timelines, possible futures, and lofi thumbnails through a Discord bot which is free for a limited time. The last selfie images from Midjourney AI were posted by zx_JB user on #show-and-tell room, Midjouney’s Discord channel.
A TikTok video of the last selfie was posted to an account called — @robotoverloards, with the bio touting “daily disturbing AI-generated images” as its mission.
One image had flashes of light behind the selfie taker, while another showed Earth levitating in the sky, suggesting that humanity has moved to a moon or a different planet. A striking aspect can be seen in the characters, some of which appear looked like aliens or zombies. Another frightening image had a figure in a black plague-type outfit with various other similar haunting figures standing behind them.
This AI-generated futuristic possibility of the end of humankind took social media by storm and resulted in various memes being generated with these wild results.
Although, the creepy predictions and ghoulish grim reapers are not all that Midjourney can do. Some users have tried a more lighthearted perspective and positive images like AI-generated cute pets.
Dall-E, a multimodal generative neural model, has moved past its novelty stage. In a recent announcement, Dall-E turned pro and marked a new stage of its life cycle. Although the tool is still in the testing phase, it matures and becomes practical with every other update.
Dell-E was created by OpenAI in January 2021, with the 12-billion GPT-3 parameter version for training the input images. The open and free tool uses standard casual masks for text prompts and sparse attention as rows/columns or convolutional attention patterns.
Almost a year later, in April 2022, the company unveiled Dall-E 2, an upgraded version of its text-to-speech generator with higher resolution and lower latency. This version was only accessible for testing by verified partners constrained in what they may submit or make with it.
In July, OpenAI began giving free access to about 1m users on the waiting list. In the latest updates, the tool will now cost US$15 for 115 credits, where each text prompt is worth 1 credit. Users also receive a few free credits to begin with, and a smaller number of freebies each month.
Dall-E is still in the novelty stage, as evidenced by a recent tweet from ketchup manufacturer Kraft Heinz Co. The tweet included a brief video that demonstrated what happened when Dall-E was given the uncomplicated command “ketchup.” The outcome was a sauce bottle that closely resembled Heinz’s goods.
We wanted to find out what A.I. thinks “ketchup” looks like. So we used A.I. to generate images of ketchup on @OpenAI. The result? Just like humans, A.I. prefers Heinz. Comment below with a ketchup-based image you want to see. #AIKetchuppic.twitter.com/NIPUx0QGWQ
It may not be artistic to pay AI to sketch a ketchup bottle. Digital illustrators like Krista Webster are concerned that enterprises who contract artists will likely accept the less expensive, computer-generated versions.
The Indian Institute of Management (IIM) Lucknow is set to open applications for an Executive Program in AI for Business starting September 4, 2022. The program is specially designed for working professionals who wish to develop their products, services, and processes with AI and data.
The students can apply for the entire online program through the official website. The AI certification program lasts 6 months and is meant for working professionals with 2 to 3 years of experience in machine learning or artificial intelligence.
The curriculum aims to equip learners with the correct knowledge about utilizing efficient technology platforms, tools, and methodologies for AI-driven business applications.
Professor Sowmya Subramaniam, Finance and Accounting, IIM Lucknow, said, “Artificial Intelligence is no longer a fringe technology for organizations. It is important for learners and professionals to upskill as per the changing market conditions and develop a thorough understanding of new technologies.”
The institute collaborated with WileyNXT to offer this online niche program to prospective students, as told by Professor VS, Prakash Attili, IT and Systems, IIM Lucknow. The enterprise will help provide the technical expertise while IIM prepares the students with sharp business acumen.
The candidates who complete the program will be awarded a certificate from IIM Lucknow.
Detect Technologies announced a collaboration with Vedanta, a leading provider of commodities dealing with zinc, silver, lead, and similar metals, to deploy T-Pulse, an AI-powered workplace safety software across the latter’s industries. Vedanta has a wide spread of contractors and employees across India, Africa, Australia, and Ireland.
Managing health, safety, and environment (HSE) for such a vast and spread-out organization can be challenging. Vedanta has been exploring AI-driven solutions to infuse security and efficiency across its network. The development of T-Pulse being piloted across Vedanta has significantly increased the lucency of workplace risks and early detections of more than 4,000 critical HSE non-compliance cases.
Sunil Duggal, CEO of Vedanta Group, said, “This partnership will further augment Vedanta’s capabilities on technology-led safety enablement. Detect Technologies’ AI and computer vision solutions will help us enhance our digital safety monitoring across all business units.”
Per Detec, T-Pulse offers a democratized and scalable solution for plug-and-play deployment. It minimizes and mitigates risks by providing actionable insights for caution-intensive work environments like construction, logistics, mining, petrochemical, pharmaceuticals, and fabrication yards.
Daniel Raj David, CEO and co-founder of Detect Technologies, said, “We appreciate the continued conviction Vedanta has shown in Detect and are excited to enable them in their journey towards improvements in ESG and safety compliance.”
In a recent announcement, DeepMind claimed that it has accurately predicted the three-dimensional structures of almost all cataloged proteins in existence. That includes more than 200 million proteins that can be found in practically anything, including people, animals, bacteria, plants, and plants. Using an AI technique called deep learning, DeepMind’s AlphaFold model can detect the 3D structure of a protein just from its 1D amino acid sequence.
Proteins which are composed of a ribbon of amino acids that folds up into a knot of intricate twists and turns, can be regarded as fundamental blocks of living beings. Because of the intrinsic flexibility of the amino acid building components, a typical protein may take on an estimated 10 to the power of 300 distinct forms. Every protein has its distinct folding configuration, so if one is altered, the protein may misfold and cease to function. Hence, understanding protein folding is highly important.
Consider a locksmith designing a key for a lock. The locksmith needs to be familiar with the structural design of the lock to be able to make the key. Now imagine the locksmith has no access to the information about the lock, they cannot create a key based on the ambiguity of the existence of the lock. Even if they successfully create one, there is no knowing it will work for the said lock. Suppose you think of medicine as a key and protein folds as a lock. In that case, you can see why researchers invest enormous time and effort decoding the folded, 3D structure of a protein they’re working with, much like the locksmith would start their key-making quest by putting together the lock’s mold. Knowing the precise structure makes it much simpler to predict where and how a molecule will bind to a particular protein as well as how that attachment can impact the protein’s folds while developing a cure.
It can take months in a lab to determine that fold—and subsequently, the function of the protein. Scientists have long experimented with automated prediction techniques like X-ray crystallography and cryo-electron microscopy to simplify the procedure. However, no method has ever come close to matching the precision attained by people. Further, they were expensive and time-consuming.
AlphaFold employs deep-learning neural networks trained on hundreds of thousands of experimentally confirmed protein structures and sequences in the PDB and other databases. When presented with a novel sequence, it initially searches databases for similar sequences that can reveal amino acids with a history of coevolving, indicating they are near in 3D space. Another method for estimating the distances between amino-acid pairs in the new sequence is to look at the structures of similar proteins that already exist.
As AlphaFold attempts to represent the 3D positions of amino acids, it iterates clues from these parallel tracks back and forth, continuously updating its estimate. It does this by using the “attention” concept to decide which amino-acid linkages are most relevant for its task at any particular time.
In December 2020, the second iteration of AlphaFold (AlphaFold2) made headlines when it won the Critical Assessment of Protein Structure Prediction (CASP) competition. The competition, which is held every two years, assesses advancement in one of biology’s most difficult problems: figuring out proteins’ three-dimensional (3D) forms only from their amino-acid sequence. In this event, the structures of the same proteins established by experimental techniques such as X-ray crystallography or cryo-electron microscopy, which fire X-rays or electron beams at proteins to build up a picture of their form, are compared to computer-software entries. After predicting structures to atomic accuracy with a median error (RMSD_95) of less than 1 Angstrom – 3 times more accurate than the next best system and comparable to experimental methods – it won CASP14 by a large margin. Further, it was acknowledged as a solution to the 50-year-old “protein-folding problem” by the organizers of CASP.
The scientific landscape had changed significantly since AlphaFold’s formal launch in July last year, when it identified about 350,000 3D proteins. To freely share this scientific information with the entire world, the Google subsidiary published and open-sourced AlphaFold one year ago and also developed the AlphaFold Protein Structure Database (AlphaFold DB). According to DeepMind, the AlphaFold DB acts as a “google search” for protein structures, giving researchers quick access to projected models of the proteins they’re researching. This allows them to concentrate their efforts and speed up experimental work. DeepMind stated that it had mapped 98.5 percent of the proteins used by the human body by the middle of 2021. It also predicted the entire ‘proteomes’ of 20 other widely studied organisms, such as mice and the bacterium Escherichia coli.
Scientists have made remarkable discoveries thanks to the AlphaFold Protein Structure database, which allowed users to see millions of protein structures. For instance, in April, Yale University researchers reviewed AlphaFold’s database to help them achieve their objective of creating a brand-new, potent malaria vaccine. And in July of last year, researchers at the University of Portsmouth employed the method to develop enzymes that will tackle pollution caused by single-use plastics. DeepMind supported World Neglected Tropical Disease Day by developing structural predictions for organisms recognized by the World Health Organization as high-priority for research, therefore advancing the study of illnesses like leprosy and schistosomiasis, which affect more than one billion people worldwide. DeepMind also plans to assist the Drugs For Neglected Diseases Initiative in the following years in identifying treatments for neglected yet widespread tropical diseases, including Chagas disease and Leishmaniasis.
Additionally, DeepMind’s publicly accessible protein structures have been included in other openly accessible databases, including Ensembl, UniProt, and OpenTargets, where millions of people use them on a daily basis.
With the recent release of predicted structures for virtually all cataloged proteins known to science in collaboration with EMBL’s European Bioinformatics Institute (EMBL-EBI), DeepMind has increased the AlphaFold DB’s size by more than 200x, from just under 1 million structures to more than 200 million structures. Researchers envision that this might significantly improve our knowledge of biology. With the inclusion of projected structures for plants, bacteria, animals, and other creatures in this release, researchers now have a wealth of new chances to utilize AlphaFold to further their study on vital topics like sustainability, food insecurity, and unrecognized illnesses.
The recent update will also result in the majority of pages on UniProt’s primary protein database having predicted structures. Additionally, all 200+ million structures will be available for mass download via Google Cloud Public Datasets, offering scientists all across the world even greater access to AlphaFold.
While it seems like Alphafold has achieved its biggest milestone, it is yet to overcome its own limitations to foster new research areas in drug discovery and the pharmaceutical industry. For instance, at present, AlphaFold cannot recognize how proteins alter in form when in contact with chemicals like medicines or other compounds that interact with proteins. Meanwhile, researchers are exploring ways to modify its training dataset and codes that will enable enhanced functionality – apart from its predictions for each amino-acid unit of a protein and associated confidence scores.
Microsoft Defender is an anti-malware component that comes with Windows PCs. With the new Threat Intelligence feature, Microsoft will utilize RiskIQ’s tech for scanning the internet and providing data to Defender’s real-time surveillance. Furthermore, RiskIQ’s data will enrich Defender’s existing dataset and provide security teams with a view of the entire attack chain.
A Microsoft executive, Vasu Jakkal, said, “Our mission is to build a safer world for all — and threat intelligence is [at] the heart of it.” Combining the services, users also get a library of raw threat intelligence and analysis from experts.
Meanwhile, the External Attack Surface Management feature will aid the security teams in understanding how an attacker views the network. It provides a way to identify all potential resources of attackers and find the unmanaged ones. Most companies that begin using a service like this are shocked by the number of internet-facing unmanaged assets they discover.
Jakkal added, “With these new tools, Microsoft is giving security teams more data to work with to protect their networks and other assets.”
A computer engineer named James Howells threw away a hard drive containing over 8,000 bitcoins (£150 million) nearly 10 years ago. Per The Guardian, Howells plans to retrieve the bitcoin hard drive using artificial intelligence to operate a mechanical arm to sort through trash. He has not given up hope and wishes to pull the job with AI and robot dogs.
Howells’ bitcoin trash story became quite popular in 2013 when he mistakenly put the wrong hard drive in the trash, losing access to his bitcoin stash, currently worth around £150 million (US$183.1 million)! He wants to convince the council of Newport in south Wales to get permission.
Robot dogs would be used as site security by the bitcoin garbage guy, as some people may recall him, to ensure that no one else would try to locate and take the hard drive. Howells also plans to employ environmental and data recovery experts in his project, which would cost about US$12.2M. The project would span over nine to twelve months, as per Howells’ own estimates.
Howells said, “One of the things we’d like to do on the actual landfill site, once we’ve cleaned it up and recovered that land, is put a power generation facility, maybe a couple of wind turbines.” He wants to create a community-owned mining facility to create bitcoin for Newport residents.
The council of Newport refused even to schedule a meeting with him to hear about his ideas. The environment is a crucial factor behind the city’s reluctance to consider his proposals.
Louis Bouchard, a Chinese scientist, has created a new free tool in collaboration with PetaPixel to restore old and deteriorated pictures with low-resolution and create slightly better ones. The free AI tool GFP-GAN (generative facial prior-generative adversarial network) merges information from 2 AI models to fill in missing details and creates the image while sustaining high quality.
Several AI technologies can create new images from inputs, but not many can fix an old picture. Conventional methods simply fine-tune existing AI models by gauging image differences. GFP-GAN uses a new approach via a pre-trained version of NVIDIA’s StyleGAN-2 model to inform the AI model at several stages of image generation.
GANs are algorithms that use 2 neural networks, the generative model to generate new examples and the adversarial model that classifies them as ‘fake’ or ‘real,’ comparing one against the other.
The restored images produced by the AI do not accurately replicate the original image. Instead, all the components that are added to replace evident traces of deterioration and brighten the original image are model predictions that introduce extra pixels.
The creators have provided a free demo for people to use the tool, along with their code to let developers implement the restoration techniques for their projects. However, the project is constrained by AI’s limitations as it guesses the missing content. The researchers believe there might be a “slight change of identity.”
Regardless of the limitations, the AI tool is doing surprisingly well in accuracy and can remove wrinkles, spots, grains, and a few other telltale signs of damage.
With the help of AI, several fields have made some amazing progress. Existing AI algorithms have greatly benefited disciplines like data analytics, large language models, and others that use enormous amounts of data to detect patterns, learn rules, and then apply them. The foundational idea behind AI is to replicate the functioning of the human brain using arithmetic and digital representations. In other words, while the human brain relies on spiking signals sent across neuron synapses, AI processes data by carrying matrix multiplications. In addition, unlike human neurons, AI models require weeks of training, consume huge power, and are powered by silicon-based chips that are currently hit with a scarcity of resources in the semiconductor industry. Therefore, scientists turned to neuromorphic computing to solve these gnawing concerns.
In essence, neuromorphic computing is the revolutionary concept of designing computer systems that can resemble the brain’s neurological structure. A neuromorphic chip like Intel’s Loihi 2 attempts to simulate the real-time, stimulus-based learning that occurs in brains. Since existing AI models are bound by computational, literal interpretations of events, it is crucial that the next generation of AI should be able to respond quickly to unusual circumstances as the human brain would. Because of how unpredictable and dynamic the world is, AI must be able to deal with any peculiar circumstances that may arise in real-time.
The emergence of neuromorphic computing has prompted major endeavors to design new, nontraditional computational systems based on recurrent neural networks, which are critical to enabling a wide range of modern technological applications such as pattern recognition and autonomous driving.
Most of the existing chip architectures adopt von Neumann architecture, which means that the network uses independent memory and processing units. Currently, data is transferred between computers by being retrieved from memory, moved to the processing unit, processed there, and then returned to memory. This constant back and forth drain both time and energy. When processing massive datasets, the bottleneck it produces is further accentuated.
Despite using less than 20 watts of electricity, human brains still beat supercomputers, proving their superior energy efficiency. By creating artificial neural systems with “neurons” (the actual nodes that process information) and “synapses” (the connections between those nodes), neuromorphic computing can replicate the function and efficiency of the brain. This AI neural network version of our neural network of synapses is called spiking neural networks (SNN), which are arranged in layers, with each spiking neuron able to fire independently and interact with the others. This allows artificial neurons to respond to inputs by initiating a series of changes. This allows researchers to alter the amount of electricity that passes between those nodes to simulate the various intensities of brain impulses.
However, this is a major setback: spiking neural networks are limited in their ability to freely select the resolution of the data they must keep or the times they access it during calculations. They can be thought of as non-linear filters that process data as it passes through them in real-time. These networks need to keep a short-term memory record of their most recent inputs to do real-time processing on a sensory input stream. Without learning, the lives of these memory traces are fundamentally constrained by the network size and the longest time scales that can be handled by the network’s parts. Therefore, developing volatile memory technologies that use fading memory traces is the need of the hour.
Since liquid environments are also necessary for biological neurons, a convergence might be reached by applying nanoscale nonlinear fluid dynamics to neuromorphic computing.
University of California San Diego researchers have created a unique paradigm in which liquids that ordinarily do not interact with light significantly on a micro- or nanoscale, support a sizable nonlinear response to optical fields. According to research published in Advanced Photonics, a nanoscale gold patch that serves as an optical heater and causes variations in the thickness of a liquid layer covering the waveguide would provide a significant light-liquid interaction effect.
Simulation result of light affecting liquid geometry, which in turn affects reflection and transmission properties of the optical mode, thus constituting a two-way light–liquid interaction mechanism. The degree of deformation serves as an optical memory allowing storing the power magnitude of the previous optical pulse and using fluid dynamics to affect the subsequent optical pulse at the same actuation region, thus constituting an architecture where memory is part of the computation process. (Image: Gao et al.)
Researchers explain that here liquid film serves as an optical memory. It operates as follows: There is a mutual interaction between the optical mode and the liquid film when a light in the waveguide modifies the geometry of the liquid surface and changes the liquid surface’s form to impact the waveguide’s optical mode’s characteristics. Notably, when the liquid geometry changes, the optical mode’s characteristics experience a nonlinear response. After the optical pulse ends, the power of the preceding optical pulse can be determined by how much the liquid layer deforms. As mentioned earlier, in contrast to conventional computing methods, the nonlinear response and the memory are located in the same spatial region, which raises the possibility of a compact (beyond von-Neumann) design in which the memory and the computational unit are housed in the same area.
The researchers show how memory and nonlinearity can be combined to create “reservoir computing,” which can carry out digital and analog tasks like handwritten image recognition and nonlinear logic gates.
Their model also makes use of the nonlocality property of liquids. Researchers can now forecast compute enhancements that are not conceivable on platforms made of solid-state materials with a finite nonlocal spatial scale. Despite nonlocality, the model falls short of contemporary solid-state optics-based reservoir computing systems. Nonetheless, the research provides a clear road map for future experimental research in neuromorphic computing seeking to test the predicted effects and investigate complex coupling mechanisms of diverse physical processes in a liquid environment for computation.
Using multiphysics simulations, the researchers predicted various unique nonlinear and nonlocal optical phenomena by investigating the interaction between light, fluid dynamics, heat transfer, and surface tension effects. They take it one step further by showing how they may be applied to construct adaptable, unconventional computational systems. Researchers propose enhancements to state-of-the-art liquid-assisted computation systems by around five orders of magnitude in space and at least two orders of magnitude in speed by using a mature silicon photonics platform.
You can check a YouTube presentation of this research here.
A team of researchers from the University of Illinois Urbana-Champaign has created a novel technique for teaching numerous agents, such as robots and drones, to cooperate using artificial intelligence. Using multi-agent reinforcement learning, a form of artificial intelligence, they created a technique for teaching numerous agents to cooperate. As it enables us to attain high degrees of coordination and collaboration across AI agents, the area of multi-agent reinforcement learning (MARL) is becoming more and more prominent. It examines how different agents interact with one another and with a shared environment, allowing us to observe how they cooperate, coordinate, compete, or collectively learn to complete an assigned assignment.
An illustration of multi-agent reinforcement learning can be a swarm of high-rise fire fighting drones attempting to stop a wildfire. To prevent the wildfire from causing more environmental damage, the drones must work together because each drone can only view a limited portion of its surroundings.
According to Huy Tran, an Illinois aerospace engineer, the study aimed to enable decentralized agent communication. The team also concentrated on circumstances where it is not immediately clear what each agent’s responsibilities or tasks should be.
Because sometimes, it can be confusing what one agent ought to do in contrast to another agent, Tran said this research experiment is far more complicated and demanding. Individual agents, such as, can cooperate and execute tasks even when communication channels are available, but what if they lack the necessary hardware or the signals are blocked, rendering communication impossible? Tran believes how agents can gradually learn to work together to complete a goal makes it an intriguing research topic.
Tran and his colleagues used machine learning to design a utility function that informs the agent when it is functioning in a way that is advantageous to the team to resolve this issue.
The team created a machine learning method that enables us to recognize when a single agent contributes to the overall team goal. This is also important cause with a swarm of robot agents accomplish common or collective goals; it can be challenging to know which agent contributed the most to make it possible. As per Tran, if you compare it to sports, one soccer player may score, but we also want to know about the teammate’s contributions, such as assists.
The researchers’ algorithms also detect whether an agent or robot is acting in a way that doesn’t align with or help achieve the goal. As a result, robot agents just opted to perform something that wasn’t helpful to the overall objective.
The research team used simulations of games like Capture the Flag and StarCraft, a well-known computer game, to evaluate their algorithms. The team was ecstatic to learn that their strategy worked well in StarCraft, which Tran noted was slightly unexpected.
According to Tran, this specific multi-agent reinforcement learning-based algorithm is relevant to many real-world scenarios, including military surveillance, robot collaboration in a warehouse, traffic signal management, delivery coordination by autonomous vehicles, and grid control.
Tran stated that Seung Hyun Kim developed most of the theory behind the proposal as a mechanical engineering undergraduate student, with Neale Van Stralen, an aerospace student, assisting with implementation. Their paper titled, “Disentangling Successor Features for Coordination in Multi-agent Reinforcement Learning,” was published in the Proceedings of the 21st International Conference on Autonomous Agents and Multi-agent Systems, which took place in May 2022.
When implemented, reinforcement learning aims to discover an optimum strategy that maximizes the anticipated reward from the surrounding environment. When reinforcement learning is used to control several agents, the term multi-agent reinforcement learning is used. Since each agent attempts to learn its strategy to maximize its reward, MARL is essentially the same as single-agent reinforcement learning. Although it is theoretically feasible to employ a single policy for all agents, doing so would require complicated communication between several agents and a central server, which is difficult in the majority of real-world situations. Instead, decentralized multi-agent reinforcement learning is utilized in reality.
It is critical in the multiple agent robot system to complete path planning in the process of avoiding interference, allocating resources, and exchanging information in a coordinated and effective manner. Most of the solutions in conventional multi-agent coordination algorithms take place in well-known settings, and the agent autonomy is constrained by the predetermined target positions and priorities for each robot or drone. To address the issue of multi-agent coordination utilizing only visual information is still insufficient. Therefore, this research study promises new avenues of multi-agent communication using multi-agent reinforcement learning.
