“DeepSeek’s R1 model thus represents a pivotal shift, suggesting that the future of gen AI lies in innovative, cost-efficient approaches rather than the traditional paradigm of scaling through sheer computational force,” Gartner researchers, including Haritha Khandabattu, Jeremy D’Hoinne, Rita Sallam, Leinar Ramos, and Arun Chandrasekaran, wrote in a research note Wednesday. Peter Rutten, research VP for performance intensive computing and worldwide infrastructure research at IDC, says the key takeaway from DeepSeek’s results is that the current approach to AI training — which is based on the theory that AI can only improve with bigger, more, and faster architecture — is not justified. “New approaches to algorithm, framework, and software for AI development deliver comparable or even better results than, for example, the latest version of ChatGPT, with the same accuracy and at a fraction of the infrastructure cost,” says Rutten. “What this means is that AI training doesn’t need to be the sole domain of hyperscalers who can afford to invest billions of dollars into large infrastructure buildouts.” source

By Andrew Levine, Winston Paes and Emily Kennedy ( February 4, 2025, 12:09 PM EST) — In March 2023, the U.S. Department of Justice’s Criminal Division introduced a three-year pilot program on compensation incentives and clawbacks.[1]… Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

Exciting developments such as DeepSeek’s R1 announcement are extending opportunities to run large language models (LLMs) on edge devices. These advancements could have profound implications for edge computing, particularly in the realms of AIOps (artificial intelligence for IT operations) and observability. By enabling real-time insights and faster automations at the edge, enterprises can enhance their operational posture, drive down costs, and improve operational efficiency and resilience. The Impact On Edge Computing Edge computing has been gaining traction to process data closer to its source, reducing latency and bandwidth usage. Edge computing technologies help firms anticipate customer needs, act on their behalf, and operate businesses efficiently in localized contexts including internet-of-things-enabled scenarios. Running LLMs on laptops and edge devices enhances these benefits by delivering powerful AI capabilities right at the edge. Training these models is a considerable challenge, something synthetic data could play a role in for AIOps, which is an approach that DeepSeek appears to have leveraged. DeepSeek-R1 claims to be as good if not better than other top-tier models, but it also offers unique advantages such as the ability to explain its answers by default. This transparency is crucial for building trust and understanding in AI-driven decisions in AIOps solutions. Processing and analyzing vast amounts of data in real time at the edge enables more responsive and intelligent edge devices. This capability is particularly valuable in scenarios when immediate decision-making is critical but connectivity to a central source or cloud resources is intermittent and unreliable. Alternative considerations are the high costs for networking and risks associated with data traveling from the edge to the cloud and data center. Some AIOps strategic objectives are to improve prediction accuracy, enhance user experiences, and produce far-reaching contextual insights for IT operations; all these stand to benefit from LLMs processing telemetry at the edge. Enhancing AIOps And Observability AIOps and observability are crucial components of modern IT operations, providing the tools needed to monitor, analyze, and optimize complex systems. Observability tools capture real-time data points, including metrics, events, logs, and traces (MELT), which are essential for understanding system behavior and performance. AIOps leverages this data to reduce alert noise, troubleshoot issues, automate remediation, and provide deep, contextual real-time insights. With LLMs running on edge devices, AIOps and observability can achieve new levels of real-time insight and automation. For instance, LLMs can analyze MELT data on the fly, identifying patterns and anomalies that might indicate potential issues, security or operational. The immediate analysis allows for quicker detection and resolution of problems, minimizing downtime and enhancing system reliability especially in environments with unreliable or irregular connectivity. The integration of smaller-footprint LLMs that can run at the edge, such as DeepSeek-R1, with AIOps can also lead to more proactive and predictive maintenance of devices and infrastructure or injection of risk-mitigating actions with no human intervention. A New Paradigm For IT Operations The integration of LLMs with edge computing and AIOps and observability represents a new paradigm for IT operations. It could be a game-changer for edge computing, AIOps, and observability if the advances of DeepSeek and others that are sure to surface run their course. This approach enables enterprises to harness the full potential of AI at the edge, driving faster and more informed decision-making. It also allows for a more agile and resilient IT infrastructure, capable of adapting to changing conditions and demands. As enterprises embrace this new paradigm, they must rethink their data center and cloud strategies. The focus will shift to a hybrid and distributed model, dynamically allocating AI workloads between edge devices, data centers, and cloud environments. This flexibility will optimize resources, reduce costs, and enhance IT capabilities, transforming data center and cloud strategies into a more distributed and agile landscape. At the center will remain observability and AIOps platforms, with the mandate for data-driven automation, autoremediation, and broad contextual insights that span the entire IT estate. Join The Conversation Register for the upcoming webinar on February 12, The Importance Of AI-Driven IT Operations And AIOps In Edge, IoT, And OT Computing. During this webinar, I will be speaking with my colleague Michele Pelino about these very topics that DeepSeek has further catapulted into the news. As always, I invite you to reach out through social media to any of us if you want to provide general feedback. If you prefer more formal or private discussions, email [email protected] to set up a meeting! You can also follow our research at Forrester.com by clicking on any of our names below. Click the names to follow our research at Forrester.com: Carlos Casanova, Michele Pelino, and Michele Goetz. source

The U.K. government has introduced its “world-first” AI Cyber Code of Practice for companies developing AI systems. The voluntary framework outlines 13 principles designed to mitigate risks such as AI-driven cyberattacks, system failures, and data vulnerabilities. The voluntary code applies to developers, system operators, and data custodians at organisations that create, deploy, or manage AI systems. AI vendors that only sell models or components fall under other relevant guidelines. “From securing AI systems against hacking and sabotage, to ensuring they are developed and deployed in a secure way, the Code will help developers build secure, innovative AI products that drive growth,” the Department for Science, Innovation, and Technology said in a press release. Recommendations include implementing AI security training programmes, developing recovery plans, carrying out risk assessments, maintaining inventories, and communicating with end-users about how their data is being used. To provide a structured overview, TechRepublic has collated the Code’s principles, who they apply to, and example recommendations in the following table. Principle Primarily applies to Example recommendation Raise awareness of AI security threats and risks System operators, developers, and data custodians Train staff on AI security risks and update training as new threats emerge. Design your AI system for security as well as functionality and performance System operators and developers Assess security risks before developing an AI system and document mitigation strategies. Evaluate the threats and manage the risks to your AI system System operators and developers Regularly evaluate AI-specific attacks like data poisoning and manage risks. Enable human responsibility for AI systems System operators and developers Ensure AI decisions are explainable and users understand their responsibilities. Identify, track, and protect your assets System operators, developers, and data custodians Maintain an inventory of AI components and secure sensitive data. Secure your infrastructure System operators and developers Restrict access to AI models and apply API security controls Secure your supply chain System operators, developers, and data custodians Conduct risk assessment before adapting models that are not well-documented or secured. Document your data, models, and prompts Developers Release cryptographic hashes for model components that are made available to other stakeholders so they can verify their authenticity. Conduct appropriate testing and evaluation System operators and developers Ensure it is not possible to reverse engineer non-public aspects of the model or training data. Communication and processes associated with end-users and affected entities System operators and developers Convey to end-users where and how their data will be used, accessed, and stored. Maintain regular security updates, patches, and mitigations System operators and developers Provide security updates and patches and notify system operators of the updates. Monitor your system’s behaviour System operators and developers Continuously analyse AI system logs for anomalies and security risks. Ensure proper data and model disposal System operators and developers Securely dispose of training data or model after transferring or sharing ownership. The Code’s publication comes just a few weeks after the government’s publication of the AI Opportunities Action Plan, outlining the 50 ways it will build out the AI sector and turn the country into a “world leader.” Nurturing AI talent formed a key part of this. Stronger cyber security measure in the U.K. The Code’s release comes just one day after the U.K.’s National Cyber Security Centre urged software vendors to eradicate so-called “unforgivable vulnerabilities,”  which are vulnerabilities with mitigations that are, for example, cheap and well-documented, and are therefore easy to implement. Ollie N, the NCSC’s head of vulnerability management, said that for decades, vendors have “prioritised ‘features’ and ‘speed to market’ at the expense of fixing vulnerabilities that can improve security at scale.” Ollie N added that tools like the Code of Practice for Software Vendors will help eradicate many vulnerabilities and ensure security is “baked into” software. More must-read AI coverage International coalition for cyber security workforce development In addition to the Code, the U.K. has launched a new International Coalition on Cyber Security Workforces, partnering with Canada, Dubai, Ghana, Japan, and Singapore. The coalition committed to work together to address the cyber security skills gap. Members of the coalition pledged to align their approaches to cyber security workforce development, adopt common terminology, share best practices and challenges, and maintain an ongoing dialogue. With women making up only a quarter of cybersecurity professionals, progress is certainly needed in this area. Why this Cyber Code matters for businesses Recent research shows that 87% of U.K. businesses are unprepared for cyber attacks, with 99% experiencing at least one cyber incident in the past year. Moreover, only 54% of U.K. IT professionals are confident in their ability to recover their company’s data after an attack. In December, the head of the NCSC warned that the U.K.’s cyber risks are “widely underestimated.” While the AI Cyber Code of Practice remains voluntary, businesses are encouraged to proactively adopt these security measures to safeguard their AI systems and reduce exposure to cyber threats. source

Classifying parties as populist Although experts generally agree that populist political leaders or parties display high levels of anti-elitism, definitions of populism vary. We use three measures to classify populist parties: anti-elite ratings from the 2019 Chapel Hill Expert Survey (CHES), Norris’ Global Party Survey and The PopuList. We define a party as populist when at least two of these three measures classify it as such. CHES, which was conducted from February to May 2020, asked 421 political scientists specializing in political parties and European integration to evaluate the 2019 positions of 277 European political parties across all European Union member states. CHES results are regularly used by academics to classify parties with regard to their left-right ideological leanings, their key party platform positions and their degree of populism, among other things. We measure anti-elitism using an average of two variables in the CHES data. First, we used “PEOPLE_VS_ELITE,” which asked the experts to measure the parties with regard to their position on direct vs. representative democracy, where 0 means that the parties support elected officeholders making the most important decisions and 10 means that “the people,” not politicians, should make the most important decisions. Second, we used “ANTIELITE_SALIENCE,” which is a measure of the salience of anti-establishment and anti-elite rhetoric for that particular party, with 0 meaning not at all salient and 10 meaning extremely salient. The average of these two measures is shown in the table below as “anti-elitism.” In all countries, we consider parties that score at or above a 7.0 as “populist.” The Global Party Survey, which was conducted from November to December 2019, asked 1,861 experts on political parties, public opinion, elections and legislative behavior to evaluate the ideological values, issue position and populist rhetoric of parties in countries on which they are an expert, classifying a total of 1,051 parties in 163 countries. We used “TYPE_POPULISM,” which categorizes populist rhetoric by parties. We added only “strongly populist” parties using this measure. In Italy, experts were asked to categorize the Center-Right coalition instead of individual parties within the coalition. The coalition includes Lega, Forza Italia and Brothers of Italy. For all three parties, we have used the coalition rating of “strongly populist.” The PopuList is an ongoing project to classify European political parties as populist, far right, far left and/or euroskeptic. The project specifically looks at parties that have “been represented in their country’s national parliament at least once” since 1989. It is based on collaboration between academic experts and journalists. The PopuList classifies parties that emphasize the will of the people against the elite as populist. This appendix uses The PopuList 3.0. Classifying parties as left, right or center We can further classify these traditional and populist parties into three groups: left, right and center. When classifying parties based on ideology, we relied on the variable “LRGEN” in the CHES dataset, which asked experts to rate the positions of each party in terms of its overall ideological stance, with 0 meaning extreme left, 5 meaning center and 10 meaning extreme right. We define left parties as those that score below 4.5 and right parties as those above 5.5. Center parties have ratings between 4.5 and 5.5. source

Cathy O’Sullivan  00:06Hello and welcome to CIO Leadership Live Australia. I’m Cathy O’Sullivan, editorial director for Foundry in Australia and New Zealand. And today, my guest is Milad Kruze, who is the Executive General Manager for IT at the Salvation Army, and he’s also one of our CIO50 alumni. Hello Milad, happy 2025 hope the new year is off to a good start for you. Milad Kruze  00:32Hello Cathy, and thanks for having me. And yes, we’re off to a flying start here in 2025. Cathy O’Sullivan  00:39Excellent, excellent. So look, one of my favorite questions to ask CIOs is about their origin story and how they got their start in tech. So can you share with us your journey and how you got your start and what have been some of the highlights of your career so far? Milad Kruze  00:56Yeah, absolutely. You know, it was about sort of 20 odd years ago now so it’s been a it’s been a while, but I came fresh out of uni, and I did a Bachelor of Information Technology, and I got right into my first formal job as a production control operator for a small company in Sydney and they did end of day processing and transactions for credit unions and banks. Sounds really exciting, right? I questioned my career options a few times as I was in that role, and it was shift work, and it was behind screens and all that sort of stuff, and I didn’t know much about the industry to go and search for other roles that I could do. And I quickly learned about sort of front facing roles and working with people. And I actually moved to Canberra and took up a service desk role, and I loved it. I love the having the ability to help people, and I sort of had this tool of trade that could do that. And I really enjoyed putting, you know, a smile on everyone’s face when I was able to help them, and that kind of started my career in it went from there and sort of went into more consulting roles, and then slowly started to manage sort of technology for small businesses. And because I was in Canberra, I naturally got into a lot of government work, and that led me to the Australian Defence Force, which ended up leading the systems engineering team for the Royal Australian Navy. And we took care of all the IT systems and all the Navy ships and ports across Australia. And that was a really cool gig. I got to exercise a few different skills there. And one day, someone asked me to do project management and lead a project. And I knew nothing about projects other than this buzzword of PRINCE2 and people going on the course. And it was a real juncture for me to figure out whether I stayed sort of in that technical world or move into management. And I decided to, yeah, sure, you know, start moving into sort of that project management realm a little bit, and that started to go to sort of more consulting areas and working for a systems integrator at the time, and the word, sort of transformation and cloud was born. So we’re very much at the forefront of working with multiple organisations across Australia, and I had the opportunity to see how well organizations did things and how other organisations didn’t. And, you know, even doing the same projects, it was easier sometimes to do it for 10,000 people than it was for 200 people. So I got real exposures to different industries, and then I decided to go out on my own, and I wanted to be the customer, because I was working with customers for so much and I’m like, well, let’s go and be the customer and do it inside organisations. And did more transformation work. Moved to Melbourne, and did some work there, around bringing organisations together, and these sort of multi year transformations, deploying productivity tools and changing the way businesses operate, and sort of landed into the CIO role. So, and I think everything through my career has kind of had a tech and mission connection there somewhere. And, you know, Tech for Good and purposeful stuff. So, yeah, it’s, it’s led me to the Salvos, and I’ve thoroughly enjoyed it. Cathy O’Sullivan  04:25So clearly helping people is something that’s a key driver for you, from those Help Desk days right through to the Defence Force, and now, as you say at the Salvation Army. And look, the Salvation Army is a household name for many, many households in Australia and New Zealand. But for anyone who’s listening who doesn’t know what the Sallies do, can you give us an overview of the organization and what you and the team are responsible for? Milad Kruze  04:52Absolutely, it’s, you know, depending on who you ask, I think you’ll get a different answer. You know, a lot of people know us for our famous Salvo stores. You know, we’ve got 400 retail stores across the country, but it’s so much broader than that. And I didn’t even know that before joining the Salvation Army, and even little over two and a half years now, I’m still learning about things that we do. We’re a Christian based organisation. We operate in about 140 countries worldwide, here in Australia, where the largest non for profit, if one of the largest for a lot of the social programs that we offer. So you know, we help everyone every 17 seconds across our social programs, and they range from anything from homelessness and housing support through to family domestic violence, disaster relief, youth services and a wide range of community services as well, such as financial counseling, etc. But we also have an aged care arm within the Salvation Army, employment services as well, and our salvo stores. And in addition to that, the other side of the

By Jeff Montgomery ( February 4, 2025, 8:58 PM EST) — Citing in part aversion to “speculative litigation,” Delaware’s Supreme Court on Tuesday reversed a ruling that kept in play potential stockholder damage claims in connection with the proposed reincorporation in Nevada of TripAdvisor and its parent, finding that business judgment deference should govern the court challenge…. Law360 is on it, so you are, too. A Law360 subscription puts you at the center of fast-moving legal issues, trends and developments so you can act with speed and confidence. Over 200 articles are published daily across more than 60 topics, industries, practice areas and jurisdictions. A Law360 subscription includes features such as Daily newsletters Expert analysis Mobile app Advanced search Judge information Real-time alerts 450K+ searchable archived articles And more! Experience Law360 today with a free 7-day trial. source

Best overall: Deel Best for global automation: Rippling Best for Growing Businesses: Papaya Global Best for Global Contractor Misclassification Protection: Oyster Best for extensibility: Remote Global payroll software providers offer tools and services to help you pay employees and contractors around the globe in their own currency, with minimal hassle, and at an affordable price. Many do so by offering employer of record (EOR) services that allow you to easily, efficiently, and compliantly hire and pay global talent without setting up an entity in each country. To help you attract and retain top talent, many providers also offer benefits packages that align with the cultural and legal expectations of the regions where you’re hiring. I evaluated and tested dozens of global payroll solutions to narrow in on the top five available to businesses today. 1 Deel Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Any Company Size Any Company Size Features Benefits Administration, Compensation Management, Direct Deposit, and more 2 Velocity Global Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Any Company Size Any Company Size Features Employee Database, Multi-Country Payroll, Payroll, and more 3 Remofirst Employees per Company Size Micro (0-49), Small (50-249), Medium (250-999), Large (1,000-4,999), Enterprise (5,000+) Any Company Size Any Company Size Features Benefits Administration, Check Printing, Compensation Management, and more Top Global Payroll Services comparison Our rating (out of 5) Starting price Number of countries Deel 4.64 stars Free 150+ Rippling 4.24 stars $8 per month per employee 185+ Papaya Global 4.27 stars $15 per employee per month 160 Oyster 4.24 stars $25 per user per month 180+ Remote 4.16 stars Free 200+ Deel: Best overall Our rating: 4.64 out of 5 Image: Deel Deel is an all-in-one payroll, human resource (HR), and human resource information system (HRIS) technology and service provider that allows you to hire employees and contractors in over 150 countries without having to build an entity in each one. On top of these, it also offers EOR and professional employer organization (PEO) services. Together, these allow you to hire employees and contractors globally with localized compliance support and get access to HR assistance and Fortune-500 employee benefits plans at affordable prices. This provider offers eight different plans starting with a free price tag to meet the needs of small to enterprise companies. Its features allow you to: Hire contractors Hire globally Process your payroll domestically and internationally Offer competitive employee benefits Manage your human resource needs at scale Engage your employees Manage employee performance Support employee learning and career development Ensure localized compliance worldwide Automate payroll and HR processes Provide job equipment to employees worldwide Access in-depth workforce, HR, and payroll reporting Outsource the management of employee immigration processes Pricing Deel offers eight plans ranging from free to $499 per month. Many of its plans use per-person pricing, meaning final prices can add up quickly based on the number of employees or contractors you onboard. Here is an overview of its plans: Deel HR: This free plan includes tools to manage your global HR workforce needs, including staying compliant with local regulations, managing time-off requests, creating people directories and organizational charts, storing signed documents, and automating HR processes. Deel US Payroll: At $19 per user per month, this plan gives you the tools to run payroll within the United States, including on-demand expert compliance support; automated W-2, W-4, and 1099 new hire filings; new hire registrations; and instant tax and wage calculations. Deel Engage: At $20 per employee per month, this plan offers tools for managing your employee performance, including learning management, career development and employee goal setting, and performance reviews. Deel Payroll: Priced at $29 per employee per month, this plan allows you to run payroll internationally with in-house payroll experts and also handles local tax filing with authorities. Deel Contractor: For $49 per month, this plan allows you to hire, manage compliance, and pay contractors in over 150 countries. Deel US PEO: Starting at $70 per employee per month, this plan offers HR, payroll, and benefits administration support in all 50 U.S. states. It also gives you access to affordable Fortune 500 benefits. Deel EOR: At $499 per month, this plan allows you to expand employee hiring to over 150 countries without having to build an entity in each location. Deel Immigration: While you must request a custom quote to learn pricing for this plan, it is designed to support hiring and relocation process for immigrant employees from over 40 countries, including the management of their visa and green card requirements. Deel pros and cons Pros Cons It offers a free HR plan with tools to manage global workforces, automate HR processes, and ensure compliance with localized insights. Some plans can be cost-prohibitive for small companies. Its PEO service gives employers access to Fortune 500 employee benefits plans at affordable prices. Some competitors, such as Oyster, allow you to hire in more than Deel’s 150 countries. Some competitors allow you to pay in more currencies than Deel does. Why I chose Deel I chose Deel as best overall because it offers something for all globally expanding businesses (and budgets). For example, its free plan gives you tools to manage all HR processes for a global workforce, from ensuring compliance with local regulations to managing time-off requests and automating processes like onboarding and offboarding employees. For larger companies looking to expand rapidly, Deel’s EOR plan enables hiring in over 150 countries without requiring you to build an entity in each, all while ensuring localized compliance support. I also like that the platform serves both companies poised to hire globally immediately and those that simply want a provider that can offer that support when they’re ready. For example, Deel offers both a US Payroll plan for domestic hiring and a separate Payroll plan with tools for international hiring. This means companies can start wherever they are, and then upgrade or add plans as their needs become more complex.

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Just a week ago — on January 20, 2025 — Chinese AI startup DeepSeek unleashed a new, open-source AI model called R1 that might have initially been mistaken for one of the ever-growing masses of nearly interchangeable rivals that have sprung up since OpenAI debuted ChatGPT (powered by its own GPT-3.5 model, initially) more than two years ago. But that quickly proved unfounded, as DeepSeek’s mobile app has in that short time rocketed up the charts of the Apple App Store in the U.S. to dethrone ChatGPT for the number one spot and caused a massive market correction as investors dumped stock in formerly hot computer chip makers such as Nvidia, whose graphics processing units (GPUs) have been in high demand for use in massive superclusters to train new AI models and serve them up to customers on an ongoing basis (a modality known as “inference.”) Venture capitalist Marc Andreessen, echoing sentiments of other tech workers, wrote on the social network X last night: “Deepseek R1 is AI’s Sputnik moment,” comparing it to the pivotal October 1957 launch of the first artificial satellite in history, Sputnik 1, by the Soviet Union, which sparked the “space race” between that country and the U.S. to dominate space travel. Sputnik’s launch galvanized the U.S. to invest heavily in research and development of spacecraft and rocketry. While it’s not a perfect analogy — heavy investment was not needed to create DeepSeek-R1, quite the contrary (more on this below) — it does seem to signify a major turning point in the global AI marketplace, as for the first time, an AI product from China has become the most popular in the world. But before we jump on the DeepSeek hype train, let’s take a step back and examine the reality. As someone who has extensively used OpenAI’s ChatGPT — on both web and mobile platforms — and followed AI advancements closely, I believe that while DeepSeek-R1’s achievements are noteworthy, it’s not time to dismiss ChatGPT or U.S. AI investments just yet. And please note, I am not being paid by OpenAI to say this — I’ve never taken money from the company and don’t plan on it. What DeepSeek-R1 does well DeepSeek-R1 is part of a new generation of large “reasoning” models that do more than answer user queries: They reflect on their own analysis while they are producing a response, attempting to catch errors before serving them to the user. And DeepSeek-R1 matches or surpasses OpenAI’s own reasoning model, o1, released in September 2024 initially only for ChatGPT Plus and Pro subscription users, in several areas. For instance, on the MATH-500 benchmark, which assesses high-school-level mathematical problem-solving, DeepSeek-R1 achieved a 97.3% accuracy rate, slightly outperforming OpenAI o1’s 96.4%. In terms of coding capabilities, DeepSeek-R1 scored 49.2% on the SWE-bench Verified benchmark, edging out OpenAI o1’s 48.9%. Moreover, financially, DeepSeek-R1 offers substantial cost savings. The model was developed with an investment of under $6 million, a fraction of the expenditure — estimated to be multiple billions —reportedly associated with training models like OpenAI’s o1. DeepSeek was essentially forced to become more efficient with scarce and older GPUs thanks to a U.S. export restriction on the tech’s sales to China. Additionally, DeepSeek provides API access at $0.14 per million tokens, significantly undercutting OpenAI’s rate of $7.50 per million tokens. DeepSeek-R1’s massive efficiency gain, cost savings and equivalent performance to the top U.S. AI model have caused Silicon Valley and the wider business community to freak out over what appears to be a complete upending of the AI market, geopolitics, and known economics of AI model training. While DeepSeek’s gains are revolutionary, the pendulum is swinging too far toward it right now There’s no denying that DeepSeek-R1’s cost-effectiveness is a significant achievement. But let’s not forget that DeepSeek itself owes much of its success to U.S. AI innovations, going back to the initial 2017 transformer architecture developed by Google AI researchers (which started the whole LLM craze). DeepSeek-R1 was trained on synthetic data questions and answers and specifically, according to the paper released by its researchers, on the supervised fine-tuned “dataset of DeepSeek-V3,” the company’s previous (non-reasoning) model, which was found to have many indicators of being generated with OpenAI’s GPT-4o model itself! It seems pretty clear-cut to say that without GPT-4o to provide this data, and without OpenAI’s own release of the first commercial reasoning model o1 back in September 2024, which created the category, DeepSeek-R1 would almost certainly not exist. Furthermore, OpenAI’s success required vast amounts of GPU resources, paving the way for breakthroughs that DeepSeek has undoubtedly benefited from. The current investor panic about U.S. chip and AI companies feels premature and overblown. ChatGPT’s vision and image generation capabilities are still hugely important and valuable in workplace and personal settings — DeepSeek-R1 doesn’t have any yet While DeepSeek-R1 has impressed with its visible “chain of thought” reasoning — a kind of stream of consciousness wherein the model displays text as it analyzes the user’s prompt and seeks to answer it — and efficiency in text- and math-based workflows, it lacks several features that make ChatGPT a more robust and versatile tool today. No image generation or vision capabilities The official DeepSeek-R1 website and mobile app do let users upload photos and file attachments. But, they can only extract text from them using optical character recognition (OCR), one of the earliest computing technologies (dating back to 1959). This pales in comparison to ChatGPT’s vision capabilities. A user can upload images without any text whatsoever and have ChatGPT analyze the image, describe it, or provide further information based on what it sees and the user’s text prompts. ChatGPT allows users to upload photos and can analyze visual material and provide detailed insights or actionable advice. For example, when I needed guidance on repairing my bike or maintaining my air conditioning unit, ChatGPT’s ability to process images

The goal, said Kramer, is to “reduce risks, security vulnerabilities, and compliance challenges tied to outdated systems. This is not just a technical shift; it’s a business enabler that allows enterprises to modernize at a pace that’s manageable. Success will also fall back on the customer adopting effective change and data management, requiring full responsibility from them. For SAP customers, the key question is not if, but how — and how quickly — they can transition.” Jeremy Roberts, senior director of research and content at Info-Tech Research Group, said, “there is a careful conflation between modernization and the cloud. We’ve known for a long time that this is the direction that major software vendors are going, but they see the cloud as a destination where their people want to be. It’s the future of SAP. Not all customers are as gung ho about the transition, however.” This offering, he said, “feels like a response to the hesitancy of major cloud customers to change their operations quickly and in significant ways. The reality is that as you grow linearly, the complexity of your organization can grow exponentially. This complexity is embodied in your ERP.” source