Category Archives: Ai

In brief AI cameras inside Amazons delivery trucks are denying drivers' bonus pay for errors they shouldnt be blamed for, it's reported.

The e-commerce giant installed the equipment in its vehicles earlier this year. The devices watch the road and the driver, and send out audio alerts if they don't like the way their humans are driving.

One man in Los Angeles told Vice that when he gets cut off by other cars, the machine would sense the other vehicle suddenly right in front of him, and squawk: Maintain safe distance! Logs of the audio alerts and camera footage are relayed back to Amazon, and it automatically decides whether drivers deserve to get bonuses or not from their performance on the road.

These workers, who are employed via contractors, claim they are unfairly denied extra pay for errors that were beyond their control or for things that dont necessarily mean theyre driving recklessly, such as tuning the radio or glancing at a side mirror.

When I get my score each week, I ask my company to tell me what I did wrong, the unnamed driver said. My [delivery company] will email Amazon and cc me, and say, Hey we have [drivers] who'd like to see the photos flagged as events, but they don't respond. There's no room for discussion around the possibility that maybe the camera's data isn't clean.

An Amazon spokesperson said alerts can be contested and are reviewed by staff at the internet giant to weed out incorrect judgments by the software.

Deepfakes arent all bad. The technology is helping trans people feel comfortable with communicating in gamer communities by changing the sound of their voice with AI algorithms.

It can be difficult for trans gamers to speak in group chats when the pitch of their voice doesnt match their gender identities; some may want to sound more feminine or masculine, typically.

A startup called Modulate is helping them generate new voices or so-called voice skins by using machine-learning software that automatically adjusts the sound of their speech. Some trans people have started testing the algorithms but havent yet used it in the wild, according to Wired.

We realized many people dont feel they can participate in online communities because their voice puts them at greater risk, Mike Pappas, Modulates CEO, said. He claimed the software only has a 15 millisecond lag when transforming someones speech in real time to a different pitch.

Early testers said they were impressed with the softwares capabilities, although Modulate declined to provide a live demo for the magazine.

The British government has promised to invest more in the AI industry and review semiconductor supply chains to make sure it has enough computational resources to support the growth of the technology.

This is how we will prepare the UK for the next ten years, and is built on three assumptions about the coming decade, the report's summary began.

"1. Invest and plan for the long-term needs of the AI ecosystem to continue our leadership as a science and AI superpower;

"2. Support the transition to an AI-enabled economy, capturing the benefits of innovation in the UK, and ensuring AI benefits all sectors and regions;

"3. Ensure the UK gets the national and international governance of AI technologies right to encourage innovation, investment, and protect the public and our fundamental values.

The first point involves funding more scholarships to help more people obtain postgraduate education in machine learning and data science. Researchers are encouraged to collaborate with others from European and US institutions.

Other parts of the plan, however, are a little bit more wishy-washy. There isnt strict actions or policies in some parts, for example, a lot of inter-agency collaboration involves formulating yet more reports to understand strategic goals in supporting the AI economy or algorithmic transparency.

Aurora, the self-driving car software biz, has started testing autonomous heavy duty Class 8 trucks capable of hauling over 14,969 kilograms with shipping giant FedEx.

The trucks will be monitored by a safety driver as they drive the 500-mile round trip from Dallas and Houston, Texas, along the I-485 interstate highway, Aurora announced this month. The company is aiming to operate fleets of fully autonomous trucks without the help of safety drivers by 2023.

You can read more about it here.

See more here:

Amazon delivery staff 'denied bonus' pay by AI cameras misjudging their driving - The Register

On 15 September 2021, the UN Office of the High Commissioner for Human Rights (OHCHR) published a report, The right to privacy in the digital age, that analyses how artificial intelligence (AI), through the use of profiling, automated decision-making and machine learning technologies, affects peoples fundamental rights and freedoms such as the right to privacy, the right to health and freedom of expression.

The OHCHR urges for a moratorium on the sale and use of AI systems that pose a serious risk to human rights and remote biometric recognition systems in public spaces until adequate safeguards are put in place. It also recommends banning AI applications that cannot ensure compliance with international human rights law.

While the report recognised that AI is instrumental in developing innovative solutions, it stressed the effects of the ubiquity of AI on peoples fundamental rights. The report looks in detail at the use of AI solutions in key public and private sectors, for example, in national security, criminal justice, employment and when managing information online.

In this respect, the OHCHR highlighted a number of risks of AI that need to be addressed by states and businesses, for example:

The report recommends addressing these risks using a comprehensive human rights-based approach and outlines possible ways to address the fundamental problems associated with AI, including the implementation of a robust legislative and regulatory framework, which prevents and mitigates any adverse effects of AI on human rights. States should ensure that any permitted interference with the right to privacy and other human rights through the use of AI does not impair the essence of these rights and is stipulated by law, pursues a legitimate purpose, is necessary and proportionate, and requires adequate justification of AI-supported decisions. The OHCHR also recommends that public and private entities systematically conduct human rights due diligence throughout the entire life cycle of the AI systems (including a human rights impact assessment), increase transparency about the use of AI and actively combat discrimination.

The press release is available here and the report is available here.

Link:

UN calls for moratorium on AI systems that pose serious risks to right to privacy and other human rights - JD Supra

The Transform Technology Summits start October 13th with Low-Code/No Code: Enabling Enterprise Agility. Register now!

A recent New York Times article concludes that new AI-powered automation tools such as Codex for software developers will not eliminate jobs but simply be a welcome aid to augment programmer productivity. This is consistent with the argument were increasingly hearing that people and AI have different strengths and there will be appropriate roles for each.

As discussed in a Harvard Business Review story: AI-based machines are fast, more accurate, and consistently rational, but they arent intuitive, emotional, or culturally sensitive. The belief is that AI plus humans is something of a centaur, greater than either one operating alone.

This idea of humans plus AI producing better outcomes has become a tenant of faith in technology. Everyone talks about humans being freed up to perform higher-level functions, but no one seems to know just what those high-level functions are, how they translate into real work and jobs, or the number of people needed to perform them.

A corollary of this augmented-workforce narrative is that not only will AI-augmented work enable people to pursue a higher level of abstract thinking, it will according to some also lift all of society to a higher standard of living. This is certainly an optimistic vision, and we can hope for that. However, this could also be a story imbued with magical thinking, with the true end-game being fully automated work.

Dont get me wrong; there is some evidence to support the view that AI will help us work rather than take our jobs. For example, AI lab DeepMind is designing new chess systems for the two intelligences to work in tandem with humans rather than opposed to them.

And Kai-Fu Lee, the Oracle of AI, also buys into this promise. In his new book, AI 2041: Ten Visions for our Future, he argues that repetitive tasks from stacking shelves to crunching data will be done by machines, freeing workers for more creative tasks. Forrester Research has likewise articulated that AI deployment enables people to better use their creative skills.

But, of course, some people are more creative than others, meaning that not everyone would benefit from AI-augmented work to the same degree. Which in turn reinforces a concern that AI-fueled automation, even in its augmented work capacity, could widen already existing income disparities.

One problem with the AI-augmented workforce promise is that it tells us AI will only take on the repetitive work we dont want to do. But not all work being outsourced to AI is routine or boring.

Look no further than the role of the semiconductor chip architect. This is a highly sophisticated profession, an advanced application of electrical engineering in arguably one of the most complex industries. If ever there was a job that might be thought of as immune from AI, this would have been a strong candidate. Yet recent advances from Google and Synopsys (among others using reinforcement learning neural network software) have shown the ability to do in hours what often required a team of engineers months to achieve.

One ever-faithful tech watcher still argued that the algorithms will optimize and accelerate time-intensive parts of the design process so that designers can focus on making crucial calls that require higher-level decision making.

More than likely, the current perception of work augmented by AI is a reflection on the current state of the technology and not an accurate view of the future when automation will be far more advanced. We first saw the potential of neural networks a decade ago, for example, and it took several years until that technology was developed to the point where it had practical advantages for consumers and business. Fueled in part by the pandemic, AI tech is now being widely implemented. Even massage therapists should take note, as a robot masseuse can now deliver a deep tissue massage. Yet, these are still early days for AI.

Caption: EMMA from AiTreat, a robot that uses artificial intelligence to deliver massages. Source: CNN

AI advances are being led by improvements in both hardware and software. The hardware side is driven by Moores Law, the idea that semiconductors improve by roughly 2x the number of transistors producing roughly equivalent performance and power efficiency gains every couple of years (and similarly drive down the costs of computing). This principle has been credited with all manner of electronic advances over the last several decades. As noted in a recent IEEE Spectrum article: The impact of Moores Law on modern life cant be overstated. We cant take a plane ride, make a call, or even turn on our dishwashers without encountering its effects. Without it, we would not have found the Higgs boson or created theInternet. Or have a supercomputer in your purse or pocket.

There are reasons to think that Moores-Law driven improvements in computing are nearing an end. But advanced engineering, ranging from chiplets to 3D chip packaging promise to keep the gains coming, at least for a while. These and other semiconductor design improvements have led one chip manufacturer to promise a 1000x performance improvement by 2025!

The expected improvements in AI software may be equally impressive. GPT-3, the third iteration of Generative Pre-trained Transformer from OpenAI, is a neural networkmodel consisting of 175 billion parameters. The system has proven capable of generatingcoherent prosefrom a text prompt. This is what it was designed to do, but it turns out that it can also generate other forms of text as well, including computer code and can also generate images. Moreover, while the belief is that AI will help people to be more creative, it could be that it is already capable of creativity on its own.

At its launch in May 2020, GPT-3 was the largest neural network ever introduced, and it remains among the largestdenseneural nets, exceeded only by Wu Dao 2.0 in China. (At 1.75 trillion parameters, Wu Dao 2.0 is another GPT-like language model and probably the most powerful neural network yet created.)

Some expectations are for GPT-4 to also grow and contain up to a trillion parameters. However, OpenAI CEO Sam Altman has said that it will not be larger than GPT-3 but will be far more efficient through enhanced data algorithms and fine tuning. Altman also alluded to a future GPT-5. The point being that neural networks have a long way to run in size and sophistication. We are indeed in the midst of an age of AI acceleration.

In the new book,Rule of the Robots: How Artificial Intelligence Will Transform Everything, author Martin Ford notes that nearly every technology startup is now, to some degree, investing in AI, and companies large and small in other industries are beginning to deploy the technology. The pace of innovation will only continue to accelerate as capital continues to pour into AI development. Clearly, whatever we are seeing now in the way of AI-powered automation, including the belief that AI will help us work rather than take our jobs, is but an early stage for whatever is still to come. As for what is coming, that remains the realm of speculative fiction.

In Burn In: A Novel of the Real Robotic Revolution, a Yale-educated lawyer is among those impacted when his firm replaced 80% of the legal staff with machine learning software. This could happen in the near future. The remaining 20% were indeed augmented by the AI, but the 80% had to find other work. In his case, he winds up doing gig work as an online personal assistant to the wealthy. Currently, startup company Yohana is working to realize a variation of this vision. The company is initially offering a blend of human and AI services, starting with a living, breathing assistantthat draws on data to tackle the to-do lists of subscribers. It will be telling to see if these assistants will be like the secretaries of yore, but wielding AI, or if they will be displaced cognitive workers.

The AI-driven transition to a largely automated world will take time, perhaps a few decades. This will bring many changes, with some being highly disruptive. Adjustments will not be easy. It is tempting to think that ultimately this will enrich the quality of human life. After all, as Aristotle said: When looms weave by themselves, mans slavery will end.But embracing the AI augmented work concept as currently articulated could blind us to the potential risks of job loss.Kate Crawford, a scholar focused on the social and political implications of technology, believes AI is the most profound story of our time and a lot of people are sleepwalking into it.

We all need to have a clear-eyed understanding of the growing potential for disruption and to prepare as best we can, largely by acquiring those skills most likely to be needed in the coming era. Companies need to do their part in providing skills training, and retraining will increasingly need to be a near continuous process as the pace of technology change accelerates. Government needs to develop public policies that direct the market forces driving automation towards positive outcomes for all, even while preparing for a growing social safety net that could include universal basic income.

Gary Grossman is the Senior VP of Technology Practice atEdelmanand Global Lead of the Edelman AI Center of Excellence.

Excerpt from:

We are sleepwalking into AI-augmented work - VentureBeat

Carolina Pinart, the lead for Nestls AI program, said intelligence was driving progress toward the corporates three key strategic objectives.

Day One of Informa Techs AI Summit in London saw Nestls global product director for new generation technologies provide an in-depth look at AIs potential to transform the multinational corporation.

Carolina Pinart, the lead for Nestls AI program, said intelligence was driving progress toward the corporates three key strategic objectives: enhanced efficiency, digitized operations and sustainability. The latter target is with a view of achieving Nestles goal of net zero carbon emissions by 2050.

If we look into Nestls progress last year, we made really solid progress on our structural savings program across all areas of manufacturing, procurement and administration, with massive opportunities for automating and AI, said Pinart.

Pinart stressed that all three initiatives had enabled Nestl to leverage AI, predictive maintenance and robotics to support factory automation and customization on the assembly line.

Nestl is also striving to expand the flow, accessibility and utility of real-time data as it is collected from operational technology networks, both in supply chain management and procurement operations.

These efforts support our drive to enhance consumer and customer centricity, which is a paramount objective for our company, as well as manufacturing, flexibility, agility and also transparency and traceability across the supply chain, Pinart explained.

Since 2019, Nestl has sought to transform its business into a cognitive enterprise, with unbiased machine learning tools that can be extensively used to help automate, respond, react and decide business outcomes.

Through machine learning, the mission statement set out by the company underscores its commitment to derive more value for its employees and customers, as well as society and its shareholders.

By 2025, Nestl aims to be fully data-driven and cognitive in its approach. It is currently working to define corporate directives for AI as part of its global program, which also covers topics such as ethics, organization, technology, education and communication.

Without the global program, we might get there eventually but certainly not at the speed that we want.

The global program runs together with multi-year strategic initiatives to automate operations and improve the experience of its workforce, through data, analytics and other technologies.

These programs leverage AI and machine learning where appropriate as a toolbox to solve business challenges, Pinart explained.

Pinart believes deployment of AI in Nestles global supply chain must start with a strategic analysis of the intended outcomes, followed by work to identify real use-cases where AI can be prioritized, deployed and scaled.

Also vital for Nestles implementation strategy was identifying the foundational data for AI to be trained on as well as other tech capabilities.

Pinart said: We are working on creating data assets and having the right platforms in place to really support the strategy and use-cases.

[Another] question is around organization and talent that varies depending on what were trying to do with AI. We dont have a single operating model but the question is, what is the right operating model to run all of this.

Finally governance, AI raises concerns on people. So both consumers and business leaders are employees as well, right? So how do we address those concerns.

View post:

Feeding The Cognitive Enterprise: Nestl Pushes AI, Predictive Maintenance and Robotics - IoT World Today

Artificial intelligence (AI) has drawn significant policy interest for quite some time now, and a federal approach is taking shape. A recent flurry of federal activity on the AI front has emanated in large part from the U.S. Department of Commerce (Commerce) including a move towards development of a risk management framework. The framework in particular may greatly influence how companies and organizations approach AI-related risks, including avoiding bias and promoting accuracy, privacy, and security.

Developing an AI Risk Management Framework

In the National Defense Authorization Act for Fiscal Year 2021 (2021 NDAA), Congress directed the National Institute of Standards and Technology (NIST) which falls under Commerce to develop a voluntary risk management framework for trustworthy [AI] systems. Following this directive, in late July, NIST issued a Request for Information (RFI) seeking input to help inform the development of what it refers to as the AI Risk Management Framework (AI RMF).

As NIST explained, the AI RMF is intended to help designers, developers, users, and evaluators of AI systems better manage risks across the AI lifecycle and aims to foster the development of innovative approaches to address characteristics of trustworthiness, including accuracy and mitigation of harmful bias. According to NIST, the development of the AI RMF will involve several iterations to encourage robust and continuing engagement and collaboration with interested stakeholders and will include open, public workshops, along with other forms of outreach and feedback.

In the RFI, NIST proposed that the AI RMF should feature eight key attributes. Specifically, NIST explained that the AI RMF should:

This directive as well as NISTs RFI outlining the initial attributes and goals of the AI RMF aligns with NISTs extensive work on similar frameworks for cybersecurity and privacy.

The RFI is the beginning of the collaborative development process at NIST with respect to the AI RMF. In the immediate near term, NIST is hosting a public workshop on October 19-21, 2021 to share feedback received in response to the RFI and discuss plans and goals for the AI RMF.

The National Artificial Intelligence Advisory Committee

On September 8, Secretary of Commerce Gina Raimondo announced that Commerce would be establishing a National Artificial Intelligence Advisory Committee (NAIAC) pursuant to a separate directive from the 2021 NDAA. The NAIAC, which will also include a subcommittee focused on the use of AI in law enforcement, will provide recommendations to the President on a broad range of AI-related topics, including the United States competitiveness in the AI space, the state of science related to AI, AI workforce issues, and opportunities for international AI cooperation.

NIST has already issued a Federal Register notice formally calling for nominations to the NAIAC. As this notice explains, the NAIAC will consist of at least nine members representing academia, industry, nonprofits, and federal laboratories who will be tasked with submitting a report to the President and Congress within one year, and then once every three years thereafter. Nominations for the NAIAC are due October 25, 2021.

Looking Ahead

These recent developments coming out of Commerce are by no means the only federal AI actions that have been taken since President Biden assumed office. For example, the Biden Administration launched AI.gov in May in order to introduce more Americans to the governments activities in the AI space. Similarly, in June, the White House Office of Science and Technology Policy and the National Science Foundation established a National Artificial Intelligence Research Resource Task Force that will help to expand access to educational tools and resources in order to spur AI innovation.

Given the responsibilities that have been assigned to it and the actions that it has already taken, Commerce and NIST in particular appears poised to be central to the federal governments approach. As such, stakeholders should consider engaging with NIST in its AI workstreams, including through participation in the upcoming workshop.

Link:

AI Risk Management Framework Is Among Emerging Federal Initiatives on AI - JD Supra

Woman at work in bright, open office workplace

Technology, and especially artificial intelligence (AI), is changing the way business gets done. In virtually every market worldwide, automation, business intelligence, and new systems are synthesizing operations, enabling teams to work smarter. While it may feel that new tech is being plugged in at an impressive pace, there are some late adopters, one of which is vendor relationship management.

Third-party vendors are essential for companies of any size. Manufacturers, suppliers and service providers are used for everything from product development to cybersecurity. At an enterprise level, the number of vendors can reach the tens of thousands. As is often the case in systems that have been around for a decade or more, interoperability is virtually non-existent, data is hard to find and efficiency is a pipe dream.

Sourcing and vendor management are more complex than ever, as companies move into emerging fields and leverage third-party suppliers to provide what cannot be internally produced. Vendors play long-term and essential roles in businesses. Many procurement departments have long held to an us versus them mentality, but as the number of necessary suppliers grows, so does the need for collaboration.

The type of vendor management in most business contexts today is sufficient for managing a handful of vendors with some success. But the demands of this are changing, and in larger companies, an entire vendor ecosystem requires more sophisticated management. In this context, the sheer volume of contacts, contracts, products, tasks and more far exceeds existing practices. Fortunately, data is what AI handles best, and new technology is being developed to get vendor relationship management up to speed.

CIOs, chief procurement leaders and financial officers regularly report on how difficult it is to understand the vendor relationship from a contract perspective, to understand how much tech a company actually owned versus used, to keep track of all of the support personnel and even to find the vendor data that lived in their ecosystem. Most of this has been and is still managed through email or on spreadsheets.

For years, client and customer communication have been managed via client relationship management (CRM) software, while vendor relationships have been left to outmoded methods. Solutions have been patchwork, piecemeal and largely ineffective. Even at the highest enterprise level, in companies with extensive technology, a problem persists: it is difficult to manage vendor relationships and assets in one place.

Brandon Card

A first reason is that there are disparate data sets, and enterprise businesses are notoriously siloed. Vendor information may reside in the legal department, the IT department and be tied to product records. A second reason this challenge has not been solved sooner is that every company hires vendors on different terms, and often each vendor on different terms. The compound challenges of these two issues alone have kept companies from making changes, but now that AI is more accessible, a smart solution is possible.

Industry leader Brandon Card is a well-known expert who has worked in development and sales for many companies, including IBM and Microsoft. He launched an application called Terzo, the first Vendor Relationship Management (VRM) system. The idea is to create a platform where vendors can now be managed in the way contacts have been managed by marketers.

CRMs already pull data from different sources, integrate with existing systems, have visibility across departments and centralize and visualize data in a meaningful way. CRMs have a long history of success in managing contexts, and the data obtained in them can trigger alerts, renewal notices and even alert marketers to churn risk. No successful company manages client relationships in any other way, and AI has already amplified the capacity of CRMs, making it easier to filter, sort and interpret data.

The same technology is at work in the emerging area of VRMs. Vendor management can be handled as data from various sources is added to the system, including contract dates, contract details, vendor supplier contact information, vendor communication records and more. Any procurement department in any company will attest to how transformational this process will be, alleviating the manual burden of record-keeping and retrieval, and leveraging the power of AI to manage these crucial relationships more effectively.

Enterprise resource planning (ERP) systems automate business processes, and are often the primary way businesses manage day-to-day operations. While these systems accomplish their purpose, they continue to be an inadequate way to manage vendors. Card explains, Across all of the systems in place today, customers rely upon ERP systems to manage all of their suppliers' data. ERP systems are a black box and cant solve vendor challenges for today, which is why people resort to spreadsheets. The problem is that, ten years ago, these companies were doing everything in-house with their own data centers.

As company offerings get more niched, and products become more technically complex, in-house solutions will not suffice. Businesses have grown and strategic vendors deliver cost-saving solutions. In most companies, entire departments cant exist without third-party vendors. AI provides the first glimmer of hope for a digital platform capable of helping businesses manage all of vendor relationships and the stakeholders that manage those relationships. These transactional relationships have now become critical partnerships, and because of AI, the technology has finally caught up to the need.

Vendor relationships directly impact a companys success. These critical partnerships require collaboration, and a vendor hub built on AI may provide that solution. It is a new way of working that complements existing systems. Easily integrated with a companys tech stack, this solution enriches data and user experiences and uncovers insights faster.

Contract data provides the single source of truth, from which AI can then extract all products, services and legal terms. Companies understand what they own and how much theyre paying for it. AI speeds up time to visibility and gets insights five to six times faster.

This is a new way of treating strategic suppliers as business partners, and empowering stakeholders with tools to more effectively connect, communicate, monitor and achieve results. As technology continues to improve and upgrade business operations, solutions like these make it possible for companies to do more.

Go here to see the original:

Vendor Management Is The New Customer Management, And AI Is Transforming The Sector Already - Forbes

A big about-face: The Clearview decision represents a major win for civil society groups and researchers focused on bringing transparency and accountability to dealings between the government and the private sector. Those groups had argued that the legal mandates were a tactic to dissuade advocacy groups from further investigating the company or others like it, and could undermine press freedom by targeting journalists and their sources.

A controversial technology: Facial recognition tools have long been under scrutiny by privacy and civil rights advocates whove raised alarm about biases with the technology generally and the disproportionate harm it can cause to marginalized communities.

After 2019 research from Open The Government shed light on police use of this surveillance including Clearviews tech and later prompted reporting on the startup, the software firm now finds itself tangled in multidistrict consumer privacy litigation in Illinois.

The debate over the subpoenas: The subpoenas included demands for communications with journalists about Clearview and its leaders, as well as information theyd uncovered about the company and its founders in public records requests.

Clearview's attorney Andrew J. Lichtman said previously that the subpoenas were served as part of its defense in the litigation in Illinois. But that case does not appear to involve Open The Government or the other parties subpoenaed, and Clearview would not explain how the groups' correspondence with journalists would be in any way relevant.

See the article here:

Clearview AI drops subpoenas of its critics - POLITICO

SAN JOSE, Calif.--(BUSINESS WIRE)--Sep 27, 2021--

Recently, MLCommons, a well-known open engineering consortium, released the results of MLPerf Inference V1.1, the leading AI benchmark suite. In the very competitive Closed Division, Inspur ranked first in 15 out of 30 tasks, making it the most successful vendor at the event.

Inspur Results in MLPerf TM Inference V1.1

Vendor

Division

System

Model

Accuracy

Score

Units

Inspur

Data CenterClosed

NF5688M6

3D-UNet

Offline, 99%

498.03

Samples/s

NF5688M6

3D-UNet

Offline, 99.9%

498.03

Samples/s

NF5488A5

DLRM

Offline, 99%

2607910

Samples/s

NF5688M6

DLRM

Server, 99%

2608410

Queries/s

NF5488A5

DLRM

Offline, 99.9%

2607910

Samples/s

NF5688M6

DLRM

Server, 99.9%

2608410

Queries/s

EdgeClosed

NE5260M5

3D-UNet

Offline, 99%

93.49

Samples/s

NE5260M5

3D-UNet

Offline, 99.9%

93.49

Samples/s

NE5260M5

Bert

Offline, 99%

5914.13

Samples/s

NF5688M6

Bert

SingleStream, 99%

1.54

Latency (ms)

NF5688M6

ResNet50

SingleStream, 99%

0.43

Latency (ms)

NE5260M5

RNNT

Offline, 99%

24446.9

Samples/s

NF5688M6

RNNT

SingleStream, 99%

18.5

Latency (ms)

NF5688M6

SSD-ResNet34

SingleStream, 99%

1.67

Latency (ms)

NF5488A5

SSD-MobileNet

SingleStream, 99%

0.25

Latency (ms)

Developed by Turing Award winner David Patterson and leading academic institutions, MLPerf is the leading industry benchmark for AI performance. Founded in 2020 and based on MLPerf benchmarks, MLCommons is an open non-profit engineering consortium dedicated to advancing standards and metrics for machine learning and AI performance. Inspur is a founding member of MLCommons, along with over 50 other leading organizations and companies from across the AI landscape.

In the MLPerf Inference V1.1 benchmark test, the Closed Division included two categories Data Center (16 tasks) and Edge (14 tasks). Under the Data Center category, six models were covered, including Image Classification (ResNet50), Medical Image Segmentation (3D-UNet), Object Detection (SSD-ResNet34), Speech Recognition (RNN-T), Natural Language Processing (BERT), and Recommendation (DLRM). A high accuracy mode (99.9%) was set for BERT, DLRM and 3D-UNET. Every model task evaluated the performance in both Server and Offline scenarios with the exception 3D-UNET, which was only evaluated in the Offline scenario. For the Edge category, the Recommendation (DLRM) model was removed and the Object Detection (SSD-MobileNet) model was added. A high accuracy mode (99.9%) was set for 3D-UNET. All models were tested for both Offline and Single Stream inference.

In the extremely competitive Closed Division, in which mainstream vendors were competing, the use of the same models and optimizers was required by all participants. Doing so provided the ability to easily evaluate and compare AI computing system performance among various vendors. Nineteen vendors including Nvidia, Intel, Inspur, Qualcomm, Alibaba, Dell, and HPE participated in the Closed Division. A total of 1,130 results were submitted, including 710 for the Data Center category, and 420 for the Edge category.

Full-Stack AI Capabilities Ramp up Performance

Inspur achieved excellent results in this MLPerf competition with its three AI servers NF5488A5, NF5688M6, and NE5260M5.

Inspur ranked first in 15 tasks covering all AI models, including Medical Image Recognition, Natural Language Processing, Image Classification, Speech Recognition, Recommendation, as well as Object Detection (SSD-ResNet34 and SSD-MobileNet). The results showcase that from Cloud to Edge, Inspur is ahead of the Industry in nearly all aspects. Inspur was able to make huge strides in performance in various tasks under the Data Center category compared to previous MLPerf events despite no changes to its server configuration. Its model performance results in Image Classification (ResNet50) and Speech Recognition (RNN-T) increased by 4.75% and 3.83% compared to the V1.0 competition just six months ago.

The outstanding performance of Inspur's AI servers in the MLPerf Benchmark Test can be credited to Inspur's exceptional system design and full-stack optimization in AI computing systems. Through precise calibration and optimization, CPU and GPU performance as well as the data communication between CPUs and GPUs were able to reach the highest levels for AI inference. Additionally, by enhancing the round-robin scheduling for multiple GPUs based on GPU topology, the performance of a single GPU or multiple GPUs can be increased nearly linearly.

Inspur NF5488A5 was the only AI server in this MLPerf competition to support eight 500W A100 GPUs with liquid cooling technology, which significantly boosted AI computing performance. Among mainstream high-end AI servers with 8 NVIDIA A100 SXM4 GPUs, Inspur's servers came out on top in all 16 tasks in the Closed Division under the Data Center category.

As a leading AI computing company, Inspur is committed to the R&D and innovation of AI computing, including both resource-based and algorithm platforms. It also works with other leading AI enterprises to promote the industrialization of AI and the development of AI-driven industries through its Meta-Brain technology ecosystem.

To view the complete results of MLPerf Inference v1.1, please visit:

Inspur Information is a leading provider of data center infrastructure, cloud computing, and AI solutions. It is the worlds 2 nd largest server manufacturer. Through engineering and innovation, Inspur delivers cutting-edge computing hardware design and extensive product offerings to address important technology arenas like open computing, cloud data center, AI, and deep learning. Performance-optimized and purpose-built, our world-class solutions empower customers to tackle specific workloads and real-world challenges. To learn more, please go to https://www.inspursystems.com/.

KEYWORD: UNITED STATES NORTH AMERICA CALIFORNIA

Read the original:

Inspur Comes Out on Top with Superior AI Performance in MLPerf Inference V1.1 - Maryville Daily Times

Do you ever think about spatial barriers?

In 1997, a chess computer named Deep Blue became the first in the world to beat a reigning chess champion. The machine, complex and high-speed, went toe-to-toe with Garry Kasparov. The AI processed information at great speed, overcoming spatial barriers with processing power.

Now, an AI like Deep Blue is less startling to humankind than it was initially. But coevolving informatics continue to reshape the frontiers of what information can do. Associate Professor Girard examines how information can re-shape our understanding of gender equality, or evaluate the resources available to the people of Afghanistan, especially when it comes to the urban-rural divide.

When it comes back down to the kind of information a brain or AI can process, there are new ways for coevolving informatics to grow. For instance, because electromagnetic signals lack mass and travel at the speed of light, because they dont have the pathway and energy constraints of other types of information. Humans are a living testament to the power of evolved information use, as we have curated a cognitive capacity beyond that of all other living creatures. If we put those understandings together, where can we go?

According to research by Florida International University, the future could mean instant interfacing of brains with cloud-based AI enabling a kind of global access that is not limited to any individual capacity for memory. This would give an intensive advantage to groups with access to that globalised information.

But how could this future work? What kind of problems and considerations would arise?

To find out more about the imminent potential of cloud-based AI and the human brain together, read more about the work of Florida International University here.

Editor's Recommended Articles

Read the rest here:

Coevolving informatics, AI and brain-computer interfacing - Open Access Government

HOUSTON, Sep. 27, 2021 /PRNewswire-PRWeb/ --Lone Star College-University Park today announced a new initiative that will enable faculty to cultivate more enriching, engaging class discussions utilizing artificial intelligence (AI). In partnership with the inquiry-based discussion platform Packback, LSC-University Park will expand faculty and student access to a platform shown to promote critical thinking, build a sense of community in class, and help improve student success.

"The power of a good class discussion is that it empowers students to engage with the curriculum beyond what's in their textbooksand help them to see that they're not alone in the questions they have or the observations they make," said Dr. Kathy Cecil-Sanchez, vice president of instruction at LSC-University Park. "That's the kind of critical thinking that Packback's platform helps to inspire, while also allowing faculty to foster deeper and more substantive engagement with their students."

Backed by entrepreneur Mark Cuban, Packback's unique platform assigns a "curiosity score" to students' discussion posts based on a range of factors, including the open-endedness of questions and the way that students reference sources. The tool also provides feedback on grammar, spelling, and offers students guidance on academic research processes like citing sources and paraphrasing. All these elements of the Packback platform are designed to spark students' intrinsic motivation, as well as to help instructors focus on providing more substantive guidance and support to students, rather than spending time on more repetitive tasks like correcting grammar.

"In my classes, Packback has been a platform that provides equity of access to the material while also pushing students to ensure that their analysis is based on facts," said Cassandra Khatri, LSC-University Park professor of political science. "Packback requires students cite the work they provide in their responses, which gets at the larger problem in political science of the gray area between analysis and takes. I really enjoy seeing my students practicing critical thinking skills as they deal with the tough issues of the day."

"I appreciate Packback's AI guidance on the writing and research, so I can focus more on the heart of the discussion topics that arise, added Jennifer Ross, LSC-University Park professor of political science. "Through Packback discussions I have seen students go from being passive learners who just accept what they read and hear, to actively taking an interest in current events and thinking about them critically."

The partnership comes in the wake of a recent research finding that inquiry-based discussion tools can spark deeper engagement and also boost academic performance. A study conducted by Packback in 2019-2020 with 10 higher education institutions, including LSC-University Park, found that students in classes that use Packback received more A, B, or C grades and fewer D, F and W grades than the control group. In the same study, students cited sources approximately 2.5 times as often in the treatment group versus the control group.

"Our work with colleges around the country is rooted in the belief that better approaches to discussion can help students build the skills that help them succeed both in and out of the classroom," said Kasey Gandham, co-founder of Packback. "As an institution that shares that vision, LSC-University Park has been an invaluable partner over the past few years, and we look forward to continuing to learn from its creative and student-centered approach in the years to come."

Packback's proprietary AI and machine learning technology provides inquiry-based online discussion to over 3,500 instructors and over 900,000 students, who have posted 22 million questions and responses to date.

Media Contact

Ben Watsky, Packback, +1 (203) 907-8930, watsky@whiteboardadvisors.com

SOURCE Packback

Continue reading here:

Lone Star College-University Park Taps AI to Inspire Curiosity in Online Discussions - inForney.com