Content provided by IBM and TNW.
Today’s AI systems are quickly evolving to become humans’ new best friend. We now have AIs that can concoct award-winning whiskey, write poetry, and help doctors perform extremely precise surgical operations. But one thing they can’t do — which is, on the surface, far simpler than all those other things — is use common sense.
Common sense is different from intelligence in that it is usually something innate and natural to humans that helps them navigate daily life, and cannot really be taught. In 1906, philosopher G. K. Chesterton wrote that “common sense is a wild thing, savage, and beyond rules.”
Robots, of course, run on algorithms that are just that: rules.
So no, robots can’t use common sense — yet. But thanks to current efforts in the field, we can now measure an AI’s core psychological reasoning ability, bringing us one step closer.
Really it comes down to the fact that common sense will make AI better at helping us solve real-world issues. Many argue that AI-driven solutions designed for complex problems, like diagnosing Covid-19 treatments for example, often fail, as the system can’t readily adapt to a real-world situation where the problems are unpredictable, vague, and not defined by rules.
Common sense includes not only social abilities and reasoning but also a “naive sense of physics.”
Injecting common sense into AI could mean big things for humans; better customer service, where a robot can actually assist a disgruntled customer beyond sending them into an endless “Choose from the following options” loop. It can make autonomous cars react better to unexpected roadway incidences. It can even help the military draw life-or-death information from intelligence.
So why haven’t scientists been able to crack the common sense code thus far?
Called the “dark matter of AI”, common sense is both crucial to AI’s future development and, thus far, elusive. Equipping computers with common sense has actually been a goal of computer science since the field’s very start; in 1958, pioneering computer scientist John McCarthy published a paper titled “Programs with common sense” which looked at how logic could be used as a method of representing information in computer memory. But we’ve not moved much closer to making it a reality since.
Common sense includes not only social abilities and reasoning but also a “naive sense of physics” — this means that we know certain things about physics without having to work through physics equations, like why you shouldn’t put a bowling ball on a slanted surface. It also includes basic knowledge of abstract things like time and space, which lets us plan, estimate, and organize. “It’s knowledge that you ought to have,” says Michael Witbrock, AI researcher at the University of Auckland.
All this means that common sense is not one precise thing, and therefore cannot be easily defined by rules.
We’ve established that common sense requires a computer to infer things based on complex, real-world situations — something that comes easily to humans, and starts to form since infancy.
Computer scientists are making (slow) but steady progress toward building AI agents that can infer mental states, predict future actions, and work with humans. But in order to see how close we actually are, we first need a rigorous benchmark for evaluating an AI’s “common sense,” or its psychological reasoning ability.
Researchers from IBM, MIT, and Harvard have created just that: AGENT, which stands for Action-Goal-Efficiency-coNstraint-uTility. After testing and validation, this benchmark is shown to be able to evaluate the core psychological reasoning ability of an AI model. This means it can actually provide a sense of social awareness and could interact with humans in real-world settings.
To demonstrate common sense, an AI model must have built-in representations of how humans plan.
So what is AGENT? AGENT is a large-scale dataset of 3D animations inspired by experiments that study cognitive development in kids. The animations depict someone interacting with different objects under different physical constraints. According to IBM:
“The videos comprise distinct trials, each of which includes one or more ‘familiarization’ videos of an agent’s typical behavior in a certain physical environment, paired with ‘test’ videos of the same agent’s behavior in a new environment, which are labeled as either ‘expected’ or ‘surprising,’ given the behavior of the agent in the corresponding familiarization videos.”
A model must then judge how surprising the agent’s behaviors in the ‘test’ videos are, based on the actions it learned in the ‘familiarization’ videos. Using the AGENT benchmark, that model is then validated against large-scale human-rating trials, where humans rated the ‘surprising’ ‘test’ videos as more surprising than the ‘expected’ test videos.
IBM’s trial shows that to demonstrate common sense, an AI model must have built-in representations of how humans plan. This means combining both a basic sense of physics and ‘cost-reward trade-offs’, which means an understanding of how humans take actions “based on utility, trading off the rewards of its goal against the costs of reaching it.”
While not yet perfect, the findings show AGENT is a promising diagnostic tool for developing and evaluating common sense in AI, something IBM is also working on. It also shows that we can utilize similar traditional developmental psychology methods to those used to teach human children how objects and ideas relate.
In the future, this could help significantly reduce the need for training in these models allowing businesses to save on computing energy, time, and money.
Robots don’t understand human consciousness yet — but with the development of benchmarking tools like AGENT, we’ll be able to measure how close we’re getting.
We all have heard not to "train for the test" when it comes to tactical fitness, and this statement is true. However, the results of a test will determine whether you get into the military as a recruit, stay in the military or advance to more selective training (special ops, Officer Candidate School (OCS), ROTC, service academy).
These tests matter significantly to your future. While it is true that training for the fitness test that includes push-ups, sit-ups, and a 1.5-mile run is only preparing you for the test and perhaps not the actual demands of a military job, there is a time when training specifically for the test is necessary.
The acronym used to describe an ideal fitness program is FITT, which stands for frequency, intensity, time and type. The secret to passing any fitness test is found in this acronym.
You must practice consistently to see results. This schedule can be every other day on a regular basis for several weeks, months or daily, depending on your preference and abilities. If you're looking to pass a fitness test, this would mean to do your testing events every other day if upper-body calisthenics and running are on the test.
If you needed to limit your overall volume of running, you could do your calisthenics, followed by a run of test length on Monday, Wednesday and Friday. On the days in between, you could mix in some non-impact cardio and leg calisthenics (squat, lunges) to aid in preparing the legs and lungs for the running event, minus the impact.
When you practice running and calisthenics for a fitness test, intensity also matters. Get used to running and doing higher-repetition calisthenics at a goal pace well above the minimum standards. Time spent doing longer, slower distances may not be the best option if you need to run a short and fast pace in order to pass a PT test.
For both the calisthenics and the cardio events of a PT test, moving with a purposeful pace is the key to your success. If you need to do 80 sit-ups in two minutes, you need to work on a pace of 20 sit-ups in 30 seconds and build up to two minutes.
The same goes for the run pace. If you need to run a seven-minute mile, you need to practice the pace of 1:45 for 400 yards and 3:30 for 800 yards to eventually do a seven-minute mile pace for the distance of your PT Test.
The amount of time each day you devote to the test is also important, but it does not have to be hours. In fact, you can block out 15 minutes of calisthenics workouts in the morning and 15 minutes of running in the afternoon if time is tight for you on any given day.
The time you spend each day (or every other day) on fitness can be molded to fit into your schedule. After all, most fitness tests in the military take from less than 15 minutes to 30 minutes of actual exertion time as they advance.
The most important factor of time is consistency and frequency of your training. You will typically need a four- to eight-week phase of training before your test for best results.
This is the key. No matter what you are testing, your workouts must be specific to the exercises being tested if you want to Improve those test scores. If it is the events of the Army Combat Fitness Test (ACFT), then get in the weight room and mix lifting (deadlift) with calisthenics, shuttle runs with weight and two-mile run workouts.
If it is simply calisthenics (push-ups, pull-ups, sit-ups or plank, and run), then you can do more with less equipment as you specifically train for the test.
Your success with this type of training also depends on your technique and how you efficiently perform the exercises. Practice is needed, but perfect practice is preferred.
You do not want to spend your entire year training specifically for a test, though the newer tactical fitness tests like the ACFT represent better models of training that force practicing strength, power, speed and agility, as well as the standard muscle stamina and cardio endurance (basic calisthenics and cardio PT tests).
Depending on your job and your future goals, your ability to get good at all the tactical elements of fitness may be required. Do not neglect training specifically for objective graded events as well as the more all-encompassing physical challenges of the day-to-day activities in the profession.
Stew Smith is a former Navy SEAL and fitness author certified as a Strength and Conditioning Specialist (CSCS) with the National Strength and Conditioning Association. Visit his Fitness eBook store if you're looking to start a workout program to create a healthy lifestyle. Send your fitness questions to email@example.com.
Whether you're thinking of joining the military, looking for fitness and basic training tips, or keeping up with military life and benefits, Military.com has you covered. Subscribe to Military.com to have military news, updates and resources delivered directly to your inbox.
© Copyright 2022 Military.com. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.
The LSAT is a test of endurance under time pressure, like a mental marathon.
It would be inadvisable to run a marathon without first training to run a full 26.2 miles. Likewise, it’s a bad idea to take the LSAT without first training with real practice tests.
That said, very few athletes run daily marathons. Instead, they vary their training with shorter intervals and complementary forms of exercise. They might focus one day on sprinting or climbing hills and another day on strength and conditioning at the gym.
In the same way, LSAT test-takers should use full practice tests judiciously. Taking one test after another, day after day, may seem impressive, but it can reinforce bad habits and lead to burnout.
Improvement comes from focused and methodical practice with careful attention to review and experimentation. Still, real practice tests belong at the core of any LSAT study strategy, as long as they’re used well.
Unlike other standardized tests, real LSAT tests are not hard to come by. In fact, the Law School Admission Council, which administers the exam, has made available more than 70 full, real, past LSAT tests for purchase, either through paperback compendiums of practice tests or through Official LSAT Prep Plus, which is currently priced at $99 and provides one year of access to an online bank of practice tests.
The LSAC also provides one free trial test online and five practice tests for members who sign up for an online account. Even more tests are available through private test prep companies.
With so many tests available, where should law school applicants start? Since the mid-1990s, practice tests have been numbered in chronological order. More exact tests provide the most relevant practice.
The LSAT has changed a bit over time. In 2007, the studying comprehension section began including a comparative passage, and in 2019 the LSAT moved to a digital format. LSATs that date back to the 1990s may include less clear questions and more elaborate types of logic games than exact tests.
It’s also easier to find discussions and explanations of questions online for more exact LSATs.
That said, sections from old LSATs can be great substitutes for experimental sections. On the actual LSAT, one section will be experimental and unscored. Experimental sections often throw test-takers for a loop, precisely because they haven’t been correctly balanced and refined. Since older tests also feel a little offbeat, they achieve the same effect.
Taking full timed practice tests is great for simulating test conditions and getting a sense of your current LSAT score range. Most of the time, however, it is better to break each practice questions into individual sections. Taking each section at full attention, separated by downtime for rest and review while the questions are fresh in your memory, is more conducive to learning than taking a full test at once.
A good LSAT study plan should start with a period of mastering fundamental techniques learned from a book, course, online program or tutor.
Once you have the basics down, practice them by taking untimed sections. Work slowly and deliberately, as if you were learning how to swim or ski for the first time. The questions you get wrong with unlimited time are exactly the kinds of questions you should focus on in your practice and review.
It may come as a surprise, but you will pick up speed more reliably through untimed practice than through timed practice. Slowly working your way through difficult questions will help you break each question into a series of steps that eventually feel intuitive and automatic, like muscle memory. In contrast, time pressure makes it too tempting to cut corners.
Once you are performing consistently with untimed practice, move to timed section practice. Periodically take full practice tests, as a marathoner might space out long-distance runs.
Weeks of timed practice will help build stamina, so you can sustain the focus you need to perform at your best. By knowing exactly what you’re up against, you’ll face less test anxiety.
Following this plan will help make test day feel like just another day of practice – hopefully your last!
Copyright 2022 U.S. News & World Report
A four-year bachelor’s degree has long been the first rung to climbing America’s corporate ladder.
But the move to prioritize skills over a college education is sweeping through some of America’s largest companies, including Google, EY, Microsoft, and Apple. Strong proponents say the shift helps circumvent a needless barrier to workplace diversity.
“I really do believe an inclusive diverse workforce is better for your company, it’s good for the business,” Ginni Rometty, former IBM CEO, told Fortune Media CEO Alan Murray during a panel last month for Connect, Fortune’s executive education community. “That’s not just altruistic.”
Under Rometty’s leadership in 2016, tech giant IBM coined the term “new collar jobs” in reference to roles that require a specific set of skills rather than a four-year degree. It’s a personal commitment for Rometty, one that hits close to home for the 40-year IBM veteran.
When Rometty was 16, her father left the family, leaving her mother, who’d never worked outside the home, suddenly in the position to provide.
“She had four children and nothing past high school, and she had to get a job to…get us out of this downward spiral,” Rometty recalled to Murray. “What I saw in that was that my mother had aptitude; she wasn’t dumb, she just didn’t have access, and that forever stayed in my mind.”
When Rometty became CEO in 2012 following the Great Recession, the U.S. unemployment rate hovered around 8%. Despite the influx of applicants, she struggled to find employees who were trained in the particular cybersecurity area she was looking for.
“I realized I couldn’t hire them, so I had to start building them,” she said.
In 2011, IBM launched a corporate social responsibility effort called the Pathways in Technology Early College High School (P-TECH) in Brooklyn. It’s since expanded to 11 states in the U.S. and 28 countries.
Through P-TECH, Rometty visited “a very poor high school in a bad neighborhood” that received the company’s support, as well as a community college where IBM was offering help with a technology-based curriculum and internships.
“Voilà! These kids could do the work. I didn’t have [applicants with] college degrees, so I learned that propensity to learn is way more important than just having a degree,” Rometty said.
Realizing the students were fully capable of the tasks that IBM needed moved Rometty to return to the drawing board when it came to IBM’s own application process and whom it was reaching. She said that at the time, 95% of job openings at IBM required a four-year degree. As of January 2021, less than half do, and the company is continuously reevaluating its roles.
For the jobs that now no longer require degrees and instead rely on skills and willingness to learn, IBM had always hired Ph.D. holders from the very best Ivy League schools, Rometty told Murray. But data shows that the degree-less hires for the same jobs performed just as well. “They were more loyal, higher retention, and many went on to get college degrees,” she said.
Rometty has since become cochair of OneTen, a civic organization committed to hiring, promoting, and advancing 1 million Black individuals without four-year degrees within the next 10 years.
If college degrees no longer become compulsory for white-collar jobs, many other qualifications—skills that couldn’t be easily taught in a boot camp, apprenticeship program, or in the first month on the job—could die off, too, University of Virginia Darden School of Business professor Sean Martin told Fortune last year.
“The companies themselves miss out on people that research suggests…might be less entitled, more culturally savvy, more desirous of being there,” Martin said. Rather than pedigree, he added, hiring managers should look for motivation.
That’s certainly the case at IBM. Once the company widened its scope, Rometty said, the propensity to learn quickly became more of an important hiring factor than just a degree.
This story was originally featured on Fortune.com
More from Fortune:
A 2007 flashback: home flippers are in trouble again
Managing Gen Z is like working with people ‘from a different country’
The Renault Nissan empire once held together by fugitive Carlos Ghosn may slowly be unraveling
PayPal tells users it will fine them $2,500 for misinformation, then backtracks immediately
Global tech major IBM, which employs over one lakh individuals in India, on Wednesday termed moonlighting an unethical practice.
Moonlighting, the practice of taking up secondary jobs after regular work hours, has recently been highlighted by many tech companies.
IBM's managing director for India and South Asia, Sandip Patel said, at the time of joining, employees sign an agreement saying they will be working only for IBM.
“…notwithstanding what people can do in the rest of their time, it is not ethically right to do that (moonlighting),” Patel told reporters on the sidelines of a company event.
Tech players in India are expressing varied opinions on moonlighting or side hustles, where an employee undertakes some other work for extra income.
Wipro's Chairman Rishad Premji had termed such behaviour by employees “cheating”.
“I share Rishad's position,” Patel said.
When asked about the company's hiring plans for India, which holds a key role as a talent base and as a market for the company, Patel said migration of employees to their hometowns during the pandemic has not completely reversed and hence, the IT industry has adopted the hybrid model of working.
Calling tier-2 and tier-3 cities as “emerging clusters”, Patel said the company has plans of deepening its presence in the country.
The company also announced that it has signed up with Airtel to offer its secured edge cloud services to the telco.
The Airtel platform backed by IBM cloud satellite will power Maruti Suzuki's initiatives to streamline play productivity and quality operations, a statement by IBM said.
IBM continues to spend millions to buy hybrid cloud companies, as the company makes its sixth acquisition in 2022 with Dialexa.
IBM continues to spend millions on buying hybrid cloud companies with the unveiling of its acquisition of engineering consulting specialist Dialexa to boost its cloud charge.
Since IBM CEO Arvind Krishna took the reins in April 2020, IBM has acquired more than 25 companies, including many hybrid cloud businesses.
In February alone, IBM acquired cloud consultant services standout Sentaca, as well as Microsoft Azure consultancy all-star Neudesic—with the two purchases squarely aimed at boosting IBM’s hybrid and multi-cloud services capabilities.
[Related: UK To Probe Amazon, Google, Microsoft’s Cloud Dominance]
Looking at the Armonk, N.Y.-based company’s purchase of Dialexa, IBM will gain 300 skilled product managers, designers, full-stack engineers and data scientists. Dialexa will become part of IBM’s Consulting business unit, which spearheads the company’s digital product engineering services in the Americas.
“Dialexa’s product engineering expertise, combined with IBM’s hybrid cloud and business transformation offerings, will help our clients turn concepts into differentiated product portfolios that accelerate growth,” said John Granger, senior vice president of IBM Consulting, in a statement.
Dialexa marks IBM’s sixth purchase in 2022 with the goal of boosting its hybrid cloud and artificial intelligence abilities.
Along with buying Dialexa, Sentaca and Neudesic, IBM has also acquired Randori, an attack surface management cybersecurity specialist that helps protect hybrid cloud environments.
Earlier this year, IBM’s CEO said hybrid cloud and artificial intelligence are top of mind for his company in terms of investment and the future.
“We are integrating technology and expertise—from IBM, our partners and even our competitors—to meet the urgent needs of our clients, who see hybrid cloud and AI as crucial sources of competitive advantage,” Krishna said in March. “And we are ready to be the catalyst of progress for our clients as they pursue the digital transformation of the world’s mission-critical businesses.”
In 2021, IBM’s hybrid cloud revenue jumped 19 percent compared with 2020, comprising 35 percent of its total revenue.
Based in Dallas and Chicago, Dialexa delivers a suite of digital product engineering services to help customers create transformative products to drive business outcomes.
Dialexa’s 300-strong engineers and skilled IT experts advise and create custom digital products for customers, which include Deere & Company, Pizza Hut U.S. and Toyota Motor North America. Financial terms of the Dialexa deal were not disclosed.
IBM said Dialexa provides deep experience delivering end-to-end digital product engineering services consisting of strategy, design, build, launch and optimization services across cloud platforms including Amazon Web Services and Microsoft Azure.
“Digital product engineering represents the tip of the spear for competitive advantage,” said Dialexa CEO Scott Harper in a statement. “IBM and Dialexa’s shared vision for delivering industry-defining digital products could be a game-changer.”
This is looking like a throw-the-baby-out-with-the-bathwater market. Take Tuesday trading: 490 of the S&P 500 stocks are lower after August inflation came in hotter than expected.
Most stocks are having a lousy year, but not all stocks have deserved the same fate. And many stocks have seen their price/earnings ratio collapse to seemingly attractive levels.
RESEARCH TRIANGLE PARK – In a move to enhance its hybrid cloud and AI capabilities, IBM will buy the digital product engineering consulting services firm Dialexa in a deal that will close later this year.
IBM announced the deal in a statement, which also notes that the purchase of the firm will “deepen IBM’s product engineering expertise and provide end-to-end digital transformation services for clients.”
When the deal closes, Dialexa will become the sixth company bought by IBM in 2022.
But Big Blue has been on a buying frenzy since April 2020, when Arvind Krishna became the company’s CEO. According to the company, IBM has acquired more than 25 other firms, with 13 to bolster IBM Consulting.
The latest acquisition of Dialexa points toward how IBM may grow its consulting services presence.
“In this digital era, clients are looking for the right mix of high-quality products to build new revenue streams and Improve topline growth,” said John Granger, senior vice president, IBM Consulting, in a statement. “Dialexa’s product engineering expertise, combined with IBM’s hybrid cloud and business transformation offerings, will help our clients turn concepts into differentiated product portfolios that accelerate growth.”
The company’s 300 employees are based in Dallas and in Chicago, and will join IBM Consulting, according to the statement. Among the firm’s clients is Toyota Motor North America, which will invest $2.5 billion in North Carolina to build the company’s first U.S. electric battery manufacturing plant in Randolph County.
Quantum computing will bring unimagined innovations to the world when it finally arrives in full glory. Still, quantum remains in the research labs at companies like IBM, Google, and Microsoft. While companies and research institutions are investing billions of dollars to increase the capacity of quantum systems, a time will come in the following years, or decades, when researchers will reach "quantum supremacy." But these large quantum marvels could also jeopardize the security of critical information systems. Researchers, including IBM are working to develop new security algorithms that will be resilient to these attacks.
While quantum can solve computing challenges far beyond what is possible today, its ability to find the factors of large prime numbers makes it the ideal cybersecurity safe cracker once quantum computing systems mature in their scale, quality, and speed. Every computer system and every bit of "secure" data could become vulnerable to attack from quantum-equipped nefarious actors. The World Economic Forum "estimate(s) that over 20 billion digital devices will need to be either upgraded or replaced in the next 10-20 years to use the new forms of quantum-resistant encrypted communication. We recommend that organizations start planning for this now.”
What constitutes "adequate size" might provide us some false comfort: a 2019 study suggested that a computer with 20 million qubits would take eight hours to break modern encryption. Today's quantum computers are on the order of only 100 qubits. But while that implies that the threat is in the distant future, one must consider that a bad actor doesn't need to wait for the massive quantum system to materialize. The "Steal now, crack later" approach leads to a latent future security threat. Consequently, organizations should deploy quantum-safe security as soon as possible to minimize future risk.
Consequently, the National Institute of Standards and Technology (NIST), a bureau of the U.S. Department of Commerce, has been conducting an ongoing search for quantum-safe security algorithms that are both secure and efficient. After all, we need our laptops, cars, and mobile phones to also be able to resist attacks from quantum-equipped bad actors. After four rounds of submissions, NIST selected four algorithms from a slate of 82 candidates. IBM Research had submitted 3 of the four chosen algorithms. All submissions have been subjected to research by industry scrutiny by government agencies, academic scientists, and mathematicians. This process is now reaching its conclusion; the NIST is expected to publish standards based on these 4 algorithms sometime in 2024.
The NIST contest covers the two aspects of security that could be vulnerable to quantum computing: public key encapsulation (used for public-key encryption and key establishment) and digital signatures (used for identity authentication and non-repudiation). For the former, NIST selected the CRYSTALS-Kyber algorithm. NIST selected three algorithms for signatures: CRYSTALS-Dilithium, FALCON, and SPHINCS+, with CRYSTALS-Dilithium as the primary algorithm in the signature category.
On September 29, GSMA announced the formation of the GSMA Post-Quantum Telco Network Taskforce, of which IBM and Vodafone are initial members, to help define policy, regulation and operator business processes to enhance protections of telecommunications in a future of advanced quantum computing. Since virtually all organizations and sectors conduct commerce on the internet, and the 800 providers whose pipes that carry all the internet traffic, the Telco industry is a good place to start. We expect other sectors to follow suit, perhaps starting with banking, government, and health care.
Given the magnitude of the potential risks, and the predominance of IBM Z systems in security-critical applications, IBM has included future-proof digital signature support in its latest z16 mainframe using CRYSTALS-Kyber and CRYSTALS -Dilithium algorithms selected by NIST. z16 implements this algorithm across multiple layers of firmware to help protect business-critical infrastructure and data from future quantum attacks. IBM has said it is also working to bring these new methods to the broader market.
In addition, IBM has developed a multi-step process to assist clients toward rapidly making institutions quantum safe. The company works with clients to identify where they are vulnerable to quantum-based cryptography attacks, assess cryptographic maturity and dependencies, and identify near-term achievable cryptographic goals and projects. The risks clients may face vary substantially based on the type of applications and data an organization handles and the state of its current cryptography.
Quantum computing's potential threat to global information security may seem to be a distant and abstract risk. However, the inevitable advances of quantum technology and the "Steal now, crack later" approach bad actors are undertaking to make quantum-safe a genuine and pressing matter for vendors and IT organizations. IBM wasted no time bringing that technology to market in the IBM z16. IBM Research has contributed three of the four algorithms the NIST quantum-safe contest has selected to be the most viable, secure, and efficient of the 70 techniques evaluated.
Beyond the NIST-approved algorithms, IBM Is working to provide “crypto agility”, helping organizations not only replace the soon-to-fail existing algorithms but also transform their security practices to remain resilient as new threats emerge in the post-quantum world. Creating crypto observability, enabling ongoing monitoring and actions on crypto-related security items, will help keep the world safer from bad actors with virtually unlimited computing capacity at their disposal.
More information can be found at here.
Disclosures: This article expresses the opinions of the authors, and is not to be taken as advice to purchase from nor invest in the companies mentioned. Cambrian AI Research is fortunate to have many, if not most, semiconductor firms as our clients, including Blaize, Cerebras, D-Matrix, Esperanto, FuriosaAI, Graphcore, GML, IBM, Intel, Mythic, NVIDIA, Qualcomm Technologies, Si-Five, SiMa.ai, Synopsys, and Tenstorrent. We have no investment positions in any of the companies mentioned in this article and do not plan to initiate any in the near future. For more information, please visit our website at https://cambrian-AI.com.