Now that it’s 2024, we can’t overlook the profound impact that Artificial Intelligence (AI) is having on our operations across businesses and market sectors. Government research has found that one in six UK organizations has embraced at least one AI technology within its workflows, and that number is expected to grow through to 2040.
With increasing AI and Generative AI (GenAI) adoption, the future of how we interact with the web hinges on our ability to harness the power of inference. Inference happens when a trained AI model uses real-time data to predict or complete a task, testing its ability to apply the knowledge gained during training. It’s the AI model’s moment of truth to show how well it can apply information from what it has learned. Whether you work in healthcare, ecommerce or technology, the ability to tap into AI insights and achieve true personalization will be crucial to customer engagement and future business success.
Inference: the Key to true personalisation
The key to personalisation lies in the strategic deployment of inference by scaling out inference clusters closer to the geographical location of the end user. This approach ensures that AI-driven predictions for inbound user requests are accurate and delivered with minimal delays and low latency. Businesses must embrace GenAI’s potential to unlock the ability to provide tailored and personalised user experiences.
Businesses that haven’t anticipated the importance of the inference cloud will get left behind in 2024. It is fair to say that 2023 was the year of AI experimentation, but the inference cloud will enable the realisation of actual outcomes with GenAI in 2024. Enterprises can unlock innovation in open-source Large Language Models (LLMs) and make true personalisation a reality with cloud inference.
Kevin Cochrane
Chief Marketing Officer at Vultr.
A new web app
Before the entrance of GenAI, the focus was on providing pre-existing content without personalization close to the end user. Now, as more companies undergo the GenAI transformation, we’ll see the emergence of inference at the edge – where compact LLMs can create personalized content according to users’ prompts.
Some businesses still lack a strong edge strategy – much less a GenAI edge strategy. They need to understand the importance of training centrally, inferring locally, and deploying globally. In this case, serving inference at the edge requires organizations to have a distributed Graphics Processing Unit (GPU) stack to train and fine-tune models against localized datasets.
Once these datasets are fine-tuned, the models are then deployed globally across data centers to comply with local data sovereignty and privacy regulations. Companies can provide a better, more personalized customer experience by integrating inference into their web applications by using this process.
GenAI requires GPU processing power, but GPUs are often out of reach for most companies due to high costs. When deploying GenAI, businesses should look to smaller, open-source LLMs rather than large hyperscale data centers to ensure flexibility, accuracy and cost efficiency. Companies can avoid complex and unnecessary services, a take-it-or-leave-it approach that limits customization, and vendor lock-in that makes it difficult to migrate workloads to other environments.
GenAI in 2024: Where we are and where we’re heading
The industry can expect a shift in the web application landscape by the end of 2024 with the emergence of the first applications powered by GenAI models.
Training AI models centrally allows for comprehensive learning from vast datasets. Centralized training ensures that models are well-equipped to understand complex patterns and nuances, providing a solid foundation for accurate predictions. Its true potential will be seen when these models are deployed globally, allowing businesses to tap into a diverse range of markets and user behaviors.
The crux lies in the local inference component. Inferring locally involves bringing the processing power closer to the end-user, a critical step in minimizing latency and optimising the user experience. As we witness the rise of edge computing, local inference aligns seamlessly with distributing computational tasks closer to where they are needed, ensuring real-time responses and improving efficiency.
This approach has significant implications for various industries, from e-commerce to healthcare. Consider if an e-commerce platform leveraged GenAI for personalized product recommendations. By inferring locally, the platform analyses user preferences in real-time, delivering tailored suggestions that resonate with their immediate needs. The same concept applies to healthcare applications, where local inference enhances diagnostic accuracy by providing rapid and precise insights into patient data.
This move towards local inference also addresses data privacy and compliance concerns. By processing data closer to the source, businesses can adhere to regulatory requirements while ensuring sensitive information remains within the geographical boundaries set out by data protection laws.
The Age of Inference has arrived
The journey towards the future of AI-driven web applications is marked by three strategies – central training, global deployment, and local inference. This approach not only enhances AI model capabilities but is vendor-agonistic, regardless of cloud computing platform or AI service provider. As we enter a new era of the digital age, businesses must recognize the pivotal role of inference in shaping the future of AI-driven web applications. While there’s a tendency to focus on training and deployment, bringing inference closer to the end-user is just as important. Their collective impact will offer unprecedented opportunities for innovation and personalization across diverse industries.
This article was produced as part of TechRadarPro’s Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro
President Biden visits the US National Institutes of Health, which under his proposed budget would receive roughly the same amount of funding in the 2025 fiscal year as in the 2023 fiscal year.Credit: Chris Kleponis/CNP/Bloomberg via Getty
US President Joe Biden today proposed modest increases in federal spending on science and innovation for the 2025 fiscal year. But that doesn’t mean his new budget will face an enthusiastic reception in Congress, which decides how much the government will spend.
Biden, a Democrat, has sought increases for many agencies in previous years but has run up against opposition among Republicans on Capitol Hill. Biden’s spending proposals for the 2024 fiscal year, which began in October, fared no better: in June 2023, after months of sparring, Democrats and Republicans agreed to spending limits for the 2024 fiscal year ― and for the 2025 fiscal year, likely quashing hopes that additional money will be poured into science.
Even after the June deal, the two sides continued to wrangle over the final numbers for the 2024 fiscal year. On 8 March, the Senate finally approved a spending package that cements the 2024 budget for most of the government’s largest science agencies. The House passed the bill on 6 March, and Biden is expected to sign it into law.
Trump versus Biden: what the rematch could mean for three key science issues
Against that backdrop, Biden’s newly published budget proposal “is nothing more than a showcase for the policies and the spending that the White House would like to pursue if it had the ability to do so, which it doesn’t,” says Michael Lubell, a physicist at the City College of New York in New York City, who tracks federal science-policy issues. “My guess is that none of this is going anywhere.”
Science advocates are already expressing dismay over some aspects of the new White House proposal. For example, the bipartisan CHIPS and Science Act, which was signed into law in 2022 to boost investments in semiconductors and science, authorized up to $35 billion in funding for science and innovation at major science agencies in the 2025 fiscal year, but the White House has requested only $20 billion, according to the American Association for the Advancement of Science (AAAS) in Washington DC. Nor has Congress followed through on those commitments.
Here are the White House’s proposed budget numbers for fiscal year 2025 for some key science-related agencies. Also noted is how each agency’s proposed funding compares to the amount appropriated for the 2024 fiscal year. The exception is for the National Institutes of Health, whose budget is compared to the amount appropriated for the 2023 fiscal year.
National Institutes of Health: $46.4 billion, 0.6% increase
The administration’s request for the National Institutes of Health (NIH) would keep the agency’s budget nearly flat for what will probably be the second year in a row. Lawmakers are still negotiating how much the NIH will receive in the 2024 fiscal year, but it is unlikely that the agency’s budget will be higher than in 2023. NIH director Monica Bertagnolli acknowledged in December that the 2024 appropriations process will be “painful”, particularly for early-career researchers. “A flat budget is a contracting budget,” she said.
In addition to the $46.4 billion the White House has requested for the agency in 2025, it has also asked for an additional $1.4 billion to support the Cancer Moonshot programme, which aims to at least halve the US cancer death rate in 25 years, and $1.5 billion for the Advanced Research Projects Agency for Health (ARPA-H), which was created in 2022 to fund high-risk, high-reward biomedical research. The White House has also requested that the Department of Health and Human Services, the parent agency of the NIH, receive $20 billion for biodefence and pandemic preparedness, of which $2.7 billion would go to the NIH.
But it is unlikely that Congress will fund these additional programmes in full, says Ellie Dehoney, the senior vice president of policy and advocacy at Research!America, a non-profit organization in Arlington, Virginia, that advocates for health research. Overall, “these are disappointing numbers”, Dehoney says. This is not “what the United States needs to stay in the lead” of biomedical research, she says.
NASA: $25.4 billion, 2% increase
Biden requested significantly less for NASA for the 2025 fiscal year than he did for the 2024 fiscal year, but his new request would still provide the agency with a little more funding than Congress appropriated. NASA’s science budget would increase by 3%, with much of that boost going to the agency’s earth science division for restructuring several planned Earth-observing missions. NASA’s planetary sciences division would receive $2.7 billion; one major uncertainty is how much of that would go towards retrieving rock samples from the Martian surface. Last year the sample-return mission was estimated to cost as much as $11 billion; NASA and the European Space Agency are now looking at whether they can reduce the price tag.
The proposed budget would slash funding for the operation of the Chandra X-Ray Observatory, a pre-eminent telescope that has been operating since 1999. The agency would also reduce funds for the operations of the Hubble Space Telescope, though much less drastically than for Chandra.
The White House is seeking a substantial boost for the US Environmental Protection Agency in the 2025 fiscal year, but Congress moved in the opposite direction last week: the agency’s overall budget in the 2024 fiscal year will be 9.6% lower than in the 2023 fiscal year. The picture is similar for the agency’s science and technology programmes, which are taking a 5.5% hit in the current fiscal year, leaving them with $758.1 million. The White House is now calling for an increase of 33.2% for those programmes in the 2025 fiscal year, which would bring the budget for science and technology to more than $1 billion.
National Science Foundation: $10.2 billion, 12% increase
Biden’s request for the National Science Foundation (NSF) is 12% above the funds appropriated for the 2024 fiscal year. The request includes $2 billion for priorities outlined in the 2022 CHIPS and Science Act, $1.4 billion for climate research and $300 million for infrastructure for large-scale research projects. The budget explicitly supports a single US extremely large telescope rather than the two such projects sought by astronomers.
The spending bill finalized last week imposed an 8.3% funding cut on the NSF — a “catastrophic” move for science, says Matt Hourihan, associate director of R&D and advanced industry at the Federation of American Scientists, an advocacy group based in Washington DC. But Biden’s request constitutes “a good budget that takes us in the right direction”, he says.
Centers for Disease Control and Prevention: $9.7 billion, 5.7% increase
The Biden administration requested $9.68 billion for the Centers for Disease Control and Prevention (CDC), the agency responsible for protecting public health. That would be a 5.7% increase over the agency’s funding for the 2023 fiscal year, but it is a smaller request than the $11.6 billion budget that the administration proposed for the 2024 fiscal year. “The request comes from, unfortunately, a return to austerity overall for discretionary funding,” says Dara Lieberman, director of government relations at Trust for America’s Health (TFAH), an advocacy group in Washington DC.
The budget includes substantial funding for efforts to modernize public health data systems: $225 million, a 28.5% increase over the amount appropriated for the 2023 fiscal year.
Department of Energy Office of Science: $8.6 billion, 4.2% increase
The Department of Energy (DOE) Office of Science, a major funder of research in the physical sciences, has weathered the budget storm better than most. The deal finalized by Congress last week increased the office’s budget for the 2024 fiscal year to more than $8.2 billion — a 1.7% increase over 2023 — and the White House is seeking another increase in the 2025 fiscal year.
The outlook is mixed for other parts of the DOE. The request for clean-energy programmes within the DOE Office of Energy Efficiency and Renewable Energy, for example, is $5.1 billion. That is more than 46% higher than the amount that Congress appropriated for the 2024 fiscal year, but 9.4% less than the amount appropriated for the 2023 fiscal year. One clear winner is the National Nuclear Security Administration, an agency within the DOE that maintains the U.S. stockpile of nuclear weapons: its budget for the 2024 fiscal year is $19.1 billion, an increase of nearly $2 billion over the 2023 fiscal year, and the White House is seeking more than $19.8 billion for the 2025 fiscal year.
Urgent question
The White House proposal sets the stage for a new round of budget negotiations, but for Carney the most pressing question is how and when Congress will resolve questions about funding the rest of the government in the current fiscal year. As it stands, much of the federal government — including the National Institutes of Health, the world’s largest public funder of biomedical research — is poised to shut down in less than two weeks unless lawmakers act. And according to the budget agreement reached between Biden and the Republicans last year, further spending cuts will kick in if the Congress doesn’t finalize the appropriations process by the end of April.
WhatsApp is currently testing a new in-app label letting you know whether or not a chat room has end-to-end encryption (E2EE).
WABetaInfo discovered the caption in the latest Android beta. According to the publication, it’ll appear underneath the contact and group name but only if the conversation is encrypted by the company’s “Signal Protocol” (Not to be confused with the Signal messaging app; the two are different.) The line is meant to serve as a “visual confirmation” informing everyone that outside forces cannot read what they’re talking about or listen to phone calls. WABetaInfo adds that the text will disappear after a few seconds, allowing the Last Seen indicator to take its place. At this moment, it’s unknown if the two lines will change back and forth or if Last Seen will permanently take the E2EE label’s place.
This may not seem like a big deal since it’s just four words with a lock icon. However, this small change is important because it indicates Meta is willing to embrace third-party interoperability.
📝 WhatsApp beta for Android 2.24.6.11: what’s new?WhatsApp is rolling out a feature to indicate when chats are end-to-end encrypted, and it’s available to some beta testers!Some users can get this feature by installing the previous updates.https://t.co/g2i5S7d9R1 pic.twitter.com/KsTa13z0BOMarch 9, 2024
See more
Third-party compatibility
On March 6, the tech giant published a report on its Engineering at Meta blog detailing how interoperability will work in Europe. The EU passed the Digital Markets Act in 2022 which, among other things, implemented new rules forcing major messaging platforms to let users communicate with third-party services.
Meta’s post gets into the weeds explaining how interoperability will work. The main takeaway is the company wants partners to use their Signal Protocol. The standard serves as the basis for E2EE on WhatsApp and Messenger, so they want everyone to be on the same playing field.
Other services don’t have to use Signal. They can use their compatible protocols, although they must demonstrate they offer “the same security guarantees”.
The wording here is pretty cut and dry: if a service doesn’t have the same level of protection, then WhatsApp won’t communicate with it. However, the beta suggests Meta is willing to be flexible. They may not completely shut out non-Signal-compliant platforms. At the very least, the company will inform its users that certain chat rooms may not be as well protected as the ones with E2EE enabled.
Interested Android owners can install the update from the Google Play Beta Program although there is a chance you may not receive the feature. WABetaInfo states it’s only available to a handful of testers. No word if WhatsApp on iOS will see the same patch.
If you’re like me, you might never have heard of continuous glucose monitors (CGMs) until the last few months, if at all. However, if your online presence has found its way on to any health or wellness algorithms, you’ve almost certainly encountered at least one or two advertisements or endorsements for the technology.
That’s because these smart glucose sensors are being touted by manufacturers and lifestyle brands as the key to unlocking and improving metabolic health, despite having been invented primarily to provide diabetes patients with real-time glucose readings.
Graph showing the worldwide search interest for CGMs increasing over the last year (Source: Google) (Image credit: Future / Canva)
But what is metabolic health, and does the information generated by these devices provide any insight that can provably help non-diabetic users? These are the questions I sought to answer when I began my journey trialing CGMs this year.
So, in addition to testing out Lingo by Abbott and the Zoe diet, I’ve spoken with nutritionists, diabetes experts and CGM gurus to find out what all the fuss is about.
What is a CGM, and who’s it for?
CGMs pack some pretty fascinating technology. Once applied, they look pretty nondescript; just a white disc stuck onto your arm and rarely larger than 1.4-inches / 3.5cm in diameter. Under the hood, however, there’s a lot going on.
CGMs insert a subcutaneous (under the skin at the layer closest to your muscle) sensor, using an algorithm to estimate blood glucose concentration based on the levels they find in the interstitial fluid (found in the spaces between cells). They do it in almost real-time, with a delay of just 15-30 minutes depending on the sensor. This data is then transmitted to a companion app and translated into a glucose reading to show how your body responds to recently eaten foods.
I spoke with Dr. Mark J. O’Connor, Assistant Professor of Medicine and practicing endocrinologist, to learn more about how these devices are used in a clinical setting.
Dr. O’Connor is an Assistant Professor of Medicine at the UMass Chan Medical School in Worcester, Massachusetts, and a practicing endocrinologist at UMass Memorial Health Care System, particularly in the UMass Diabetes Center of Excellence. He’s also conducted a small research study on the utility of using CGMs to help diabetes patients in emergency departments.
He explains: “A CGM is a useful tool for people with diabetes, because it gives them a lot more insight into their blood sugar levels, and it gives us as healthcare providers more insight so that we can make the best recommendations.
“It also helps people make better decisions about exercise, food, and other factors that affect blood sugar levels. For many people with diabetes, it’s a Godsend.”
The technology has come a long way since 1999, when it gained Food and Drug Association approval for clinical use in the US. Abbott, one of three US companies manufacturing the majority of CGMs worldwide, is one of the first to take its technology into the consumer space; its consumer offering, Lingo, arrived in the UK in January, and pending FDA clearance is set to be released in the US later this year.
I spoke with Olivier Ropars, Division VP of Lingo Biowearables, about the company’s expansion into direct-to-consumer CGMs.
Olivier Ropars
Ropars joined Abbott in 2022 to run Lingo, Abbott’s consumer Biowearables division, where he is currently overseeing the growth of the product as it comes to market in various regions. He previously held positions at eBay, StubHub and McKinsey.
The key benefit to direct-to-consumer CGMs, says Ropars, lies in understanding your metabolic health – which, broadly speaking, is the absence of metabolic disorders, including conditions such as type 1 diabetes, obesity, and inflammatory bowel diseases. They can also be vital tools in measuring the long-term effects of hypoglycemia (low blood glucose) and hyperglycemia (high glucose levels).
As Ropar explains, a series of studies conducted by Abbott employees and consultants found post-meal glucose spikes correlate to sleep and mental health difficulties and increased hunger. More worryingly, they were also associated with the development of type 2 diabetes and an “increased risk of seven out of the 10 leading causes of death in the US; that’s expected to be the same in the UK.” These include cardiovascular disease, liver failure, kidney failure, and increased risk of developing Alzheimer’s.
(Image credit: Shutterstock / Dragoljub Bankovic)
CGMs: Fad or fact
“The current healthcare system is entirely focused on curing diseases or trying to catch diseases early. There’s very little focus on disease prevention,” says Ropars, citing findings that health trackers such as pedometers have an “overwhelmingly positive” impact on people’s lifestyles. CGMs, he argues, will have the same effect.
Ropars says three groups of people have been particularly receptive to Lingo. The first is people, particularly athletes, looking to optimize their health and performance. These bio-hackers are a core demographic across many CGM offerings; the recently shuttered Supersapiens CGM-based platform (which also used Abbott sensors) enticed users with the promise they’d “never bonk again”, referring to the exercise-induced hypoglycemia endurance athletes can experience when they haven’t eaten enough carbohydrates. Indeed, early studies have shown CGMs might be useful for athletes in determining ideal carbohydrate intake – but as of yet, this isn’t proven.
The second and third groups are more general; people wanting to resolve age- or health-related issues who haven’t seen success in one-size-fits-all solutions; and those with a family history of disease wanting to monitor health and take preventative measures.
Many brands, such as Zoe in the UK and Super Sapiens in the US use Abbott’s CGMs paired with their own proprietary platforms for data analysis and lifestyle guidance. (Image credit: Shutterstock / Jeniffer Fontan)
O’Connor can likewise see the benefits: “To me, it does make sense that giving people access to more information would be helpful. To see the effect on your blood sugar of routine exercise or a healthy diet is positive reinforcement to continue to make healthy behavioral changes.” A small study funded by Dexcom, another major CGM manufacturer, supports this, seeing a small group of people generally improve their lifestyle after using a CGM.
Lauren Johnson Reynolds, AKA the London Wellness Coach, is a homeopath, nutritionist, and health coach with a special interest in hormone balance following her own experience with polycystic ovarian syndrome (PCOS). Early studies have shown a correlation between glucose tolerance, diabetes, and PCOS, which is why Reynolds tried out CGMs after giving birth led her HBA1-C to elevate to a near pre-diabetic range.
“There were certain foods and combinations that I felt wouldn’t cause any issues, but were spiking my blood sugar quite badly,” she says. “It was very eye-opening, and I became very aware of how stress impacted my blood sugar.”
Lauren Johnson Reynolds
Following a series of health difficulties relating to stress and PCOS, Reynolds turned to homeopathic solutions to relieve her symptoms, beginning a journey of healing and self-understanding that culminated in her studying an integrated Homeopathy and Nutritional therapy course. Now, she shares her knowledge, passion and holistic approach to wellness as a health coach.
A double-edged needle
While there’s plenty of buzz around the potential for CGMs, there’s perhaps not significant enough proof to convince some practitioners. Katherine Metzelaar, MSN, RDN, CD, founder and CEO of Bravespace Nutrition, is one of many concerned parties. “There’s this general desire to want to know what’s going on inside of our bodies – but it’s not as simple as that.
“We have yet to hear about any positive impacts of continuous glucose monitoring outside of the management of diabetes, and I think they cause more harm than good.”
Katherine Metzelaar, MSN, RDN, CD
Passionate about helping people battling disordered eating, Katherine Metzelaar’s approach is to help cultivate a positive food and body relationship by creating a weight-inclusive space and educating clients on the impacts of diet culture. Her specialism is in eating disorder recovery, body image concerns, and weight-inclusive nutrition therapy, which she practices using the HAES (health at every size) approach.
Metzelaar specializes in disordered eating and food positivity, and fears CGMs are another chapter in a legacy of illusory diet and wellness fads that contribute to an ineffective approach to health.
“It could lead to oversimplifying nutrition, increased stress and anxiety around food, and unnecessary restriction or inaccurate and inadequate advice around what to do with that information,” she says. “My recommendation is usually not to do it. Even for clients I work with that have diabetes, it leads to them feeling anxious when they see their glucose spike.”
She’s also broadly unconvinced by the usefulness of the data provided, especially the difficulty in parsing what specifically might have elevated glucose levels when eating different foods.
Despite her own positive experience, Reynolds has similar concerns about the broader use of CGMs. “I recommend them to a few clients but not the majority… When you step back and look at it, it’s quite extreme, but we’re so used to monitoring our sleep and our recovery, it’s become very normalized.”
Lingo by Abbott provides real-time insight into blood glucose levels, presenting spikes as “Lingo counts” and encouraging users to stay below a target each day. The target is based on your health data; factors such as age, gender and weight. (Image credit: Future)
“Modern society has lost track of hunger signals and thirst signals. We’re told to drink eight glasses of water a day and eat three meals and two snacks, and so that’s what we do. It’s taking away from that mind-body connection. So, although I think they are extremely useful for people who have blood-sugar issues, have a condition like PCOS, or want to try CGMs for a short amount of time, for the average person I don’t think it’s necessary.”
Metzelaar also has concerns about the self-led nature of many CGM platforms. “To my understanding, there’s no oversight from a doctor or a dietitian to interpret that information and add clarity to it.” This, she says, could lead users to change their behaviors based on what could simply be the normal curve of what happens during digestion.
I tried two different CGMs as part of my research for this article; Zoe Health and Lingo. While I don’t have any metabolic health diagnoses, I’ve had a lot of digestive and stress-related issues in recent months – but for me, this technology is still perhaps a little extreme. (Image credit: Future)
Knowledge is power – until it’s not
As of right now, there are numerous consumer-based studies underway, including large-scale efforts being conducted by CGM platforms including Levels and Signos. In the UK, Zoe Health combines gut microbiome, blood sample, and bowel movement data with its CGM readings, the results of which can be discussed with a Zoe nutritionist. Clearly, there’s plenty of confidence from these providers that there’s something in this CGM craze worth fighting for.
That’s certainly the resounding take of Ropars, who says we’re only just beginning to unlock the potential of CGMs: “It’s not like buying a diet book; that may not work for everyone. A CGM provides personalized information, so you can make the adjustments that matter the most for you.”
So, are CGMs a fad or the future of wellbeing? As of right now, they don’t definitively support non-diabetics to improve their lifestyle, nor can studies fully demonstrate the biomarkers they track have a measurable impact on long-term health and wellbeing.
Still, the appetite for the technology and the research being conducted into metabolic health, blood glucose, and the relationship with long-term health means, for better or worse, we’re likely going to be seeing a lot more about CGMs in the coming years.
Tenstorrent, the firm led by legendary chip architect Jim Keller, the mastermind behind AMD‘s Zen architecture and Tesla’s original self-driving chip, has launched its first hardware. Grayskull is a RISC-V alternative to GPUs that is designed to be easier to program and scale, and reportedly excels at handling run-time sparsity and conditional computation.
Off the back of this, Tenstorrent has also unveiled its Grayskull-powered DevKits – the standard Grayskull e75 and the more powerful Grayskull e150. Both are inference-only hardware designed for AI development, and come with TT-Buda and TT-Metalium software. The former is for running models right away, while the latter is for users who want to customize their models or write new ones.
The Santa Clara-based tech firm’s milestone launch comes hot on the heels of a partnership with Japan’s Leading-edge Semiconductor Technology Center (LSTC). Tenstorrent’s RISC-V and Chiplet IP will be used to build a state-of-the-art 2nm AI Accelerator, with the ultimate goal of revolutionizing AI performance in Japan.
By the power of Grayskull!
The Grayskull e75 model is a low-profile, half-length PCIe Gen 4 board with a single Grayskull processor, operating at 75W. The more advanced e150 model is a standard height, 3/4 length PCIe Gen 4 board containing one Grayskull processor operating at up to 200W, and balancing power and throughput.
Tenstorrent processors comprise a grid of cores known as Tensix Cores and come with network communication hardware so they can talk with one another directly over networks, instead of through DRAM.
The Grayskull DevKits support a wide range of models, including BERT for natural language processing tasks, ResNet for image recognition, Whisper for speech recognition and translation, YOLOv5 for real-time object detection, and U-Net for image segmentation.
The Grayskull e75 and e150 DevKits are available for purchase now at $599 and $799, respectively.
In the rapidly evolving world of technology, Microsoft has taken a bold step forward, with CEO Satya Nadella at the helm, charting a course for the company’s future in artificial intelligence (AI). Microsoft, a titan in the tech industry, has recently achieved a remarkable feat by overtaking Apple in market value, claiming the title of the world’s most valuable public company. This achievement is a testament to Microsoft’s relentless pursuit of innovation and its unwavering commitment to enhancing the capabilities of individuals and organizations across the globe.
Under Nadella’s leadership, Microsoft is not just riding the wave of AI; it is steering it. The company’s strategy is deeply rooted in the integration of AI into its vast array of products and services. A shining example of this is Microsoft’s collaboration with OpenAI, which has led to the creation of the Copilot tool. This innovative tool is designed to amplify productivity and foster creativity, positioning Microsoft as a leader in the AI domain. The company’s focus is not limited to technological advancements; it also encompasses the broader implications of AI, including its impact on political and social stability in the United States.
Microsoft CEO Nadella on the future of AI in 2024
The transformative potential of AI is immense, with the power to redefine user interfaces and reasoning engines, thereby revolutionizing entire software categories. Microsoft’s vision extends beyond the current smartphone era, with an eye on the future of AI-driven devices and interfaces. This forward-thinking approach suggests that Microsoft is not only interested in software but is also keen on pioneering hardware innovations to maintain its edge in the competitive tech landscape.
Here are some other articles you may find of interest on the subject of Microsoft and artificial intelligence :
Nadella’s personal interests, such as his fondness for cricket and its increasing popularity in the U.S., offer a glimpse into the intersection of global cultural trends and technology. This perspective underscores the CEO’s broader understanding of the role cultural dynamics play in shaping the tech industry. As Microsoft navigates the complex terrain of international business, particularly in markets like China, the company remains dedicated to supporting multinational enterprises and tapping into the global talent pool. Upholding intellectual property rights and adhering to regulatory standards are at the forefront of Microsoft’s agenda, especially in light of significant collaborations, such as the one with OpenAI.
Looking ahead to 2024, Microsoft, under Nadella’s strategic direction, is poised to fully embrace the flourishing AI landscape. The company is focused on developing innovative AI features that will not only empower users but also redefine the technological horizon. Microsoft’s commitment to growth extends to areas of personal interest and societal importance, signaling a new era of tech that is both advanced and attuned to the needs of the world.
Filed Under: Technology News, Top News
Latest timeswonderful Deals
Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, timeswonderful may earn an affiliate commission. Learn about our Disclosure Policy.
The field of artificial intelligence (AI) is undergoing a significant transformation, and the Turing Institute is at the forefront of this exciting era. Named after the legendary AI pioneer Alan Turing, the institute has become a beacon of innovation, turning theoretical concepts into practical applications that are beginning to reshape our world.
Since the mid-2000s, AI has experienced a surge in growth, driven by breakthroughs in machine learning. The effectiveness of these systems is largely dependent on the quality of the training data they receive. This process, known as supervised learning, allows AI to learn from examples. One of the most critical developments in this area has been the creation of neural networks. These networks, inspired by the human brain, enable machines to process and interpret vast amounts of data.
The future of generative AI
Among the most notable advancements in AI is the creation of sophisticated language models, such as GPT-3. These models have the ability to generate text that is so similar to human writing that it can be difficult to distinguish between the two. The versatility of these models is remarkable, and they are being used in a variety of applications. However, they are not without their flaws. These AI systems can sometimes produce errors, demonstrate biases, and raise concerns about issues such as toxicity and compliance with laws like the General Data Protection Regulation (GDPR).
Despite the impressive capabilities of current AI systems, they still fall short in certain areas. For instance, AI does not yet fully understand context, nor does it possess consciousness or reasoning abilities. This distinction highlights the gap between what AI can do and the full spectrum of human intelligence, which encompasses more than just language skills and pattern recognition.
The pursuit of General AI, which aims to replicate the full range of human intellectual abilities, raises profound philosophical and ethical questions. As AI-generated content becomes more prevalent online, we must consider the responsibilities associated with this content and the potential impact of AI on society, including the feedback loops it may create.
To address some of these challenges, researchers are exploring new approaches that combine symbolic AI, which operates based on a set of rules, with the data-driven methods used by large AI systems. This combination is expected to yield more robust and capable AI technologies. Additionally, the development of multimodal AI, which can process and understand various types of data such as text, images, and videos, is set to expand the possibilities of what AI can achieve.
The Turing Institute is playing a critical role in pushing the boundaries of AI while also addressing the ethical considerations that accompany these technological advances. As AI continues to progress, the goal is not to replace human capabilities but to augment them, creating tools that enhance our abilities and contribute positively to society. The future of generative AI is not only about technological innovation but also about navigating the complex landscape of societal implications that come with it.
Image Credit: Turing Institute
Filed Under: Guides, Top News
Latest timeswonderful Deals
Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, timeswonderful may earn an affiliate commission. Learn about our Disclosure Policy.
In a recent series of lectures at the Turing Institute, the UK’s hub for data science and artificial intelligence research, experts gathered to discuss the future of generative AI. This technology, which can independently generate new content, has been making significant strides, from crafting written text to creating visual art and complex legal documents. The series culminated with a session that paid homage to Alan Turing, a pioneer in computing, and delved into the exciting trajectory of generative AI.
Artificial intelligence has been on a remarkable journey, marked by steady progress and recent breakthroughs in machine learning that have propelled AI capabilities forward. Central to these advancements is the training data that AI systems use to learn. Typically, AI is fed labeled data in a process known as supervised learning, which it then uses to predict outcomes or generate new content.
During the lecture, the focus was on neural networks, which are complex structures modeled after the human brain. These networks are crucial for AI’s ability to recognize patterns and make decisions. The speakers highlighted how the combination of large data sets, affordable computing power, and scientific discoveries in deep learning have expanded what AI can achieve.
The future of generative AI
Here are some other articles you may find of interest on the subject of generative AI :
One of the most significant milestones in AI was the introduction of the Transformer architecture and large language models like GPT-3. These have greatly improved AI’s text generation capabilities, making it more realistic and pushing the boundaries of machine creativity. However, the lecture also pointed out the challenges that come with such powerful technology, including errors, biases, toxicity, and copyright issues, as well as the need to comply with GDPR to protect privacy and data.
AI’s understanding is still not perfect and is often constrained by the contexts of its training data. This limitation has led to philosophical and ethical discussions about the potential for general artificial intelligence and the concept of machine consciousness. The Turing Test, a historical benchmark for AI’s ability to mimic human intelligence, was reconsidered in light of these new developments.
Looking to the future, the lecture suggested that AI could soon develop multimodal capabilities, combining text, images, sound, and video to create personalized content. This could change the way we interact with technology, making it more intuitive and responsive to individual preferences.
The Turing Lecture series has shed light on the significant impact of generative AI and the ethical considerations and limitations that come with its use and development. As AI continues to advance, it is poised to redefine content creation and many other areas, leading to a future where it may become increasingly difficult to tell apart content created by humans from that generated by machines.
Filed Under: Technology News, Top News
Latest timeswonderful Deals
Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, timeswonderful may earn an affiliate commission. Learn about our Disclosure Policy.
Jaron Lanier, a pioneer of virtual reality and a key figure in the early days of the internet, has evolved into a critical voice within the technology sector. Despite his contributions to the industry, including founding companies that were acquired by tech giants like Google, Adobe, and Oracle, and currently working at Microsoft, Lanier expresses concerns about the direction technology, particularly AI, has taken. He emphasizes the need for a more ethical and transparent approach to AI development and usage.
Lanier’s background is deeply rooted in the tech world. He has founded companies that have caught the attention of giants like Google, Adobe, and Oracle. Now, as part of Microsoft, he brings his wealth of knowledge to one of the most influential tech companies on the planet. Despite his insider status, Lanier maintains a critical stance, constantly evaluating the industry’s trajectory and its impact on society.
When it comes to AI, Lanier sees it as a collective human achievement rather than a standalone marvel. He warns of the dangers associated with AI services that are free but rely on advertising revenue, which can manipulate user behavior. Lanier’s message is clear: the tech industry must adopt a more ethical and transparent approach to AI, ensuring that it benefits society as a whole.
The future of AI with Jaron Lanier
Here are some other articles you may find of interest on the subject of artificial intelligence :
The debate over open-source AI models is heating up. Lanier recognizes their ability to drive innovation but also points out the significant threat they pose in disseminating false information. He urges the tech community to handle these models with caution and responsibility, to prevent the spread of harmful disinformation.
One of Lanier’s key concerns is data provenance. He believes that individuals should be fairly compensated for their data, which is used to train AI systems. He introduced the idea of “data dignity,” advocating for people to receive fair payment for their contributions to the data economy. This concept is crucial in ensuring that the benefits of AI and data collection are shared with those who provide the raw material—the users themselves.
The emergence of deepfakes—highly realistic and manipulated videos or audio recordings—poses a new challenge in the realm of digital authenticity. Lanier stresses the importance of data provenance in combating these sophisticated forgeries. Establishing the origin and authenticity of data is essential in the fight against digital deception.
Lanier also calls for regulation in the AI industry. He believes that tech leaders should be more receptive to regulatory frameworks that ensure AI is developed and used in an ethical manner. However, he acknowledges the complex relationship between tech firms and policymakers, which could complicate the push for regulation.
Free speech within tech companies is another value Lanier holds dear. He argues that fostering open dialogue can lead to better outcomes for the industry and society at large. By encouraging a culture of open communication, companies can harness diverse perspectives to drive innovation and address ethical concerns.
Despite his critical views, Lanier remains dedicated to the tech industry. He strikes a balance between his roles as an innovator and a critic, setting a personal threshold for his engagement. His critiques are not meant to undermine the industry but to steer it towards a more responsible and conscientious path.
Lanier’s stance in the tech industry is one of caution and responsibility. As a virtual reality pioneer and a key figure at Microsoft, he challenges us to reconsider our relationship with technology. His calls for ethical AI, data dignity, and regulatory oversight reflect a deep concern for the direction of the industry. Lanier’s insights encourage us to think about the broader consequences of technological advancements and to strive for a future where innovation benefits everyone. His voice is not just a warning; it’s a call to action for a more thoughtful and inclusive tech landscape.
Filed Under: Gadgets News
Latest timeswonderful Deals
Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, timeswonderful may earn an affiliate commission. Learn about our Disclosure Policy.
In an era where digital transformation is at the heart of business evolution, Generative AI stands out as a marvel of innovation. This subset of artificial intelligence, known for its ability to create original content—from text to images—holds immense potential. However, its capabilities also make it a double-edged sword in the realm of cybersecurity. For businesses leveraging this powerful technology, the need for a robust security framework is non-negotiable. This article delves into the methods and strategies to secure AI business models against the burgeoning threats in cyberspace.
Securing Data for AI Models
At the core of generative AI’s power is data—vast quantities of it. The integrity and security of this data are paramount. It begins with data discovery and classification, a meticulous process that sorts through the digital deluge to identify sensitive information and safeguard it appropriately. Cryptography then comes into play, transforming this data into a cipher that is impenetrable without the correct keys, effectively nullifying data breaches.
In tandem with these measures, access controls are indispensable. With techniques like Multifactor Authentication (MFA), the security net tightens, ensuring that only those with verified credentials can reach the sensitive nucleus of AI models.
How to protect your AI Models
Once the data is locked down, attention shifts to the AI models themselves. These models are the engines of generative AI, and like any engine, they can be tampered with or corrupted. Routine scans for malicious Cybersecurity code, fortifying systems against cyber attacks, and role-based access control are critical defense strategies. Moreover, the pedigree of the data sources feeding into these models must be scrutinized for trustworthiness and legality, with APIs serving as secure conduits for data flow and interaction.
Other articles you may find of interest on the subject of Generative AI :
Deploying generative AI is not a set-and-forget affair. It demands continuous vigilance from Cybersecurity threats. Monitoring the inputs that feed the AI is crucial to prevent the propagation of misinformation or malicious content. Semantic guardrails can curtail the misuse of AI-generated content, while machine learning tools adeptly detect and respond to anomalies and threats.
Supporting these efforts are SIEM systems, which act as the watchtowers of cybersecurity, providing real-time alerts and insights. Complementing SIEM, SOAR solutions automate the responses to low-level security events, ensuring rapid containment and resolution.
AI Cybersecurity
The IT infrastructure is the backbone that supports the lofty ambitions of generative AI. It must embody the CIA Triad—confidentiality, integrity, and availability. Each component of the infrastructure, from the humblest server to the most complex network, must be treated as a potential vulnerability and fortified accordingly.
No AI system operates in a vacuum. Governance and compliance are the ethical compass and legal scaffolding that ensure AI operates within the boundaries of moral and legal acceptability. Establishing governance frameworks that dictate how AI should be used and ensuring compliance with ever-evolving regulatory landscapes are as vital as any technical safeguard.
In summary, the security of generative AI is a multi-faceted challenge that extends from the granular level of data protection to the broader strokes of governance and compliance. It requires a harmonious blend of advanced technologies, stringent policies, and continuous monitoring. The key to harnessing the full potential of generative AI lies in constructing a security architecture that is as dynamic and intelligent as the AI it seeks to protect.
Businesses must recognize that securing generative AI is not a hurdle but an enabler of innovation. In doing so, they not only defend against the specters of cyber threats but also build the trust and reliability that are the currency of the digital economy. With the right security framework, the promise of generative AI can be fully realized, propelling businesses towards a future where creativity and cybersecurity go hand in hand.
Filed Under: Guides, Toys
Latest timeswonderful Deals
Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, timeswonderful may earn an affiliate commission. Learn about our Disclosure Policy.
