 |
|
|
|||||
Welcome to AI Decoded, Fast Companys weekly newsletter that breaks down the most important news in the world of AI. You can sign up to receive this newsletter every week here. Exclusive Interview With Ciscos Jeetu Patel Data centers are popping up all over the world to support the quickly growing demand for all kinds of AI apps and services. Cisco, of course, is no stranger to the data center, and its been working hard over the past few years to make itself a vital part of the AI technology stack. I asked Cisco EVP and chief product officer Jeetu Patel how he sees the current situation in generative AI, and about how his company fits into the picture. The interview has been edited for length and clarity. Can you give me your 30,000-foot view of the transition to generative AI? Were now moving from this mode of chatbots intelligently answering questions for us to us now moving into a mode where agents are conducting tasks and jobs almost fully autonomously on behalf of humans. As that happens, theres going to be an augmentation of the capacity of billions and billions of agents thatll actually get added on over the course of the next few years. But the requirements that you have around low-latency, high-performance, high-energy-efficiency infrastructure, as well as around safety and security so that the user can establish trust with these AI systems, is going to have to be fully reimagined. Can you describe in simple terms how Cisco plays in the AI tech stack? At the very baseline, we build our own silicon and ASICs [application-specific integrated circuits] for the network itself. I think were the only non-Nvidia silicon provider that is part of Nvidias reference architecture where our networking is tied with their GPUs and we actually make sure that those work together in a reference architecture that an enterprise can deploy. We then have our own systems, which are the physical boxes for the networks and the servers on the compute side, and the optics and the optical systems that actually can do ultra-long haul data center interconnect, as well as interconnect between clusters. We then provide the safety and security platform thats needed to secure AI as wellwere one of the largest security players in the market. We provide a data platform in Splunk. Were actually building our own bespoke custom models for security and networking. You mentioned latency as a key challenge. How critical is response time for AI applications? If it takes three seconds for an AI voice agent to respond to you, you know its a robot and you dont want to talk to it. But if you do it within 500 milliseconds, you have a very different kind of behavior from the human. In our user testing, outside of efficacy, latency is one of the most important things. It has to be interruptible and it has to have enough training on EQ [emotional intelligence] and sentiment analysis, so that if youre sounding annoyed, it doesnt say, Hows the weather today? How do you handle the security challenges with multiple AI models? Most of these models are putting their own safety and security guardrails in the models. But models can get tricked through jailbreaking techniques. Weve built a product that not only does the visibility of what data is flowing through the model and when the model is getting fine-tuned, so you can do a continuous validation. . . . We validate the model within a matter of minutes through an algorithmic red-teaming exercise rather than it taking weeks or months for companies to validate the model. We jailbroke DeepSeek within 48 hours. We can take that model and then create runtime enforcement guardrails for every application developer. The end outcome is that no developer has to rebuild the security stack every time they build an application, and no model provider needs to be responsible for every single way that a model can be jailbroken. So every app developer building on top of DeepSeek will benefit from this pool of knowledge that Cisco knows about how to jailbreak the model and how to protect against that? Thats exactly right. We believe that you need a neutral party that provides a common substrate of security for every app developer, every model builder, every agent developer, so that the developer can innovate fearlessly. Are AI companies putting big data centers in the Middle East because they have plenty of power and room to grow, or is it to better service customers in that region? Its literally both. You dont have enough power to fuel all the demand for AI right now. The amount of usage that OpenAI is getting right now is literally like breaking the internet. They came up with $20 a usertheyre losing money on $20 a user, from what the industry says. So they added a plan for $200 a user. My guess is theyre going to lose money at $200 a user. They have a plan for $2,000 a user. They will lose money for $2,000 a user. Tha’s not a bad thing. It tells you that there is intrinsic demand. The demand for data centers is going to be insatiable for a very long time. As models get more efficient over time, youll have small models with very large context windowsyou might have a million-token context window, very small model, very small data set with a very small footprint to be able to get the inference done. But were not quite there yet. Is it because of inference costs that they cant make money? Whats the big cost driver? Right now its the usage and the cost of GPUs. Its expensive. But the beauty about this is its the wrong thing to focus on to get a company to profitability at this stage. What they should focus on is the acquisition of as many users as possible so that they can have the daily workflow fusion of ChatGPT for both consumers and enterprises. Once that happens, they can figure out a way to optimize later. But right now, starting to optimize would be putting cycles in the wrong thing. There are two dimensions that are a really good signal for the AI industry: Nvidia is wildly profitable, and ChatGPT is losing money. Both those things tell you the future of AItheres such insatiable demand that even at a very high cost, people are buying GPUs and theyre willing to lose money because the adoption is so high. More AI coverage from Fast Company: Why OpenAI and Microsofts AI partnership might be headed for a breakup AI is supercharging war. Could it also help broker peace? AI tools collect your data across devices. Here’s how to be selective about what you share The debate over state-level AI bans misses the point Want exclusive reporting and trend analysis on technology, business innovation, future of work, and design? Sign up for Fast Company Premium.
Category:
E-Commerce
Research on misinformation and disinformation has become the latest casualty of the Trump administrations restructuring of federal research priorities. Following President Donald Trumps executive order on ending federal censorship, the National Science Foundation canceled hundreds of grants that supported research on misinformation and disinformation. Misinformation refers to misleading narratives shared by people unaware that content is false. Disinformation is deliberately generated and shared misleading content, when the sharer knows the narrative is suspect. The overwhelming majority of Americans95%believe misinformations misleading narratives are a problem. Americans also believe that consumers, the government and social media companies need to do something about it. Defunding research on misinformation and disinformation is, thus, the opposite of what Americans want. Without research, the ability to combat misleading narratives will be impaired. The attack on misleading narrative research Trumps executive order claims that the Biden administration used research on misleading narratives to limit social media companies free speech. The Supreme Court had already rejected this claim in a 2024 case. Still, Trump and GOP politicians continue to demand disinformation researchers defend themselves, including in the March 2025 censorship industrial complex hearings, which explored alleged government censorship under the Biden administration. The U.S. State Department, additionally, is soliciting all communications between government offices and disinformation researchers for evidence of censorship. Trumps executive order to restore free speech, the hearings and the State Department decision all imply that those conducting misleading narrative research are enemies of the First Amendments guarantee of free speech. These actions have already led to significant problemsdeath threats and harassment includedfor disinformation researchers, particularly women. So lets tackle what research on misinformation and disinformation is and isnt. Misleading content Misinformation and disinformation researchers examine the sources of misleading content. They also study the spread of that content. And they investigate ways to reduce its harmful impacts. For instance, as a social psychologist who studies disinformation and misinformation, I examine the nature of misleading content. I study and then share information about the manipulation tactics used by people who spread disinformation to influence others. My aim is to better inform the public about how to protect themselves from deception. Sharing this information is free speech, not barring free speech. Yet, some think this research leads to censorship when platforms choose to use the knowledge to label or remove suspect content or ban its primary spreaders. Thats what U.S. Rep. Jim Jordan argued in launching investigations in 2023 into disinformation research. It is important to note, however, that the constitutional definition of censorship establishes that only the governmentnot citizens or businessescan be censors. So private companies have the right to make their own decisions about the content they put on their platforms. Trumps own platform, Truth Social, bans certain material such as sexual content and explicit language, but also anything moderators deem as trying to trick, defraud, or mislead us and other users. Yet, 75% of the conspiracy theories shared on the platform come from Trumps account. Further, both Trump and Elon Musk, self-proclaimed free-speech advocates, have been accused of squelching content on their platforms that is critical of them. Musk claimed the suppression of accounts on X was a result of the sites algorithm reducing the reach of a user if theyre frequently blocked or muted by other, credible users. Truth Social representatives claim accounts were banned due to bot mitigation procedures, and authentic accounts may be reinstated if their classification as inauthentic was invalid. Is it censorship? Republicans say social media companies have been biased against their content, censoring it or banning conservatives unfairly. The censorship industrial complex hearings held by the House Foreign Affairs South and Central Asia Subcommittee were based on the premise that not only was misleading narrative research part of the alleged censorship industrial complex, but that it was focused on conservative voices. But there isnt evidence to support this assertion. Research from 2020 shows that conservative voices are amplified on social media networks. When research does show that conservative authors have posts labeled or removed, or that their accounts are suspended at higher rates than liberal content, it also reveals that it is because conservative posts are significantly more likely to share misinformation than liberal posts. This was found in a recent study of X users. Researchers tracked whose posts got tagged as false or misleading more in community notesXs alternative and Metas proposed alternative to fact checkingand it was conservative posts, because they were more likely to include false content than liberal posts. Furthermore, an April 2025 study shows conservatives are more susceptible to misleading content and more likely to be targeted by it than liberals. Misleading America Those accusing misleading narrative researchers of censorship misrepresent the nature and intent of the research and researchers. And they are using disinformation tactics to do so. Heres how. The misleading information about censorship and bias has been repeated so much through the media and from political leaders, as evident in Trumps executive order, that many Republicans believe its true. This repetition produces what psychologists call the illusory truth effect, where as few as three repetitions convince the human mind something is true. Researchers have also identified a tactic known as accusation in a mirror. Thats when someone falsely accuses ones perceived opponents of conducting, plotting or desiring to commit the same transgressions that one plans to commit or is already committing. So censorship accusations from an administration that is removing books from libraries, erasing history from monuments and websites, and deleting data archives constitute accusations in a mirror. Other tactics include accusation by anecdote. When strong evidence is in short supply, people who spread disinformation point repeatedly to individual stories (sometimes completely fabricated) that are exceptions to, and not representative of, the larger reality. Facts on fact-checking Similar anecdotal attacks are used to try to dismiss fact-checkers, whose conclusions can identify and discredit disinformation, leading to its tagging or removal from social media. This is done by highlighting an incident where fact-checkers got it wrong. These attacks on fact-checking come despite the fact that many of those most controversial decisions were made by platforms, not fact-checkers. Indeed, fact-checking does work to reduce the transmission of misleading content. In studies of the perceived effectiveness of professional fact-checkers versus algorithms and everyday users, fact-checkers are rated the most effective. When Republicans do report distrust of fact-checkers, its because they perceive the fact-checkers are biased. Yet research shows little bias in choice of who is fact-checked, just that prominent and prolific speakers get checked more. When shown fact-checking results of specific posts, even conservatives often agree the right decision was made. Seeking solutions Account bans or threats of account suspensions may be more effective than fact-checks at stopping the flow of misinformation, but they are also more controversial. They are considered more akin to censorship than fact-check labels. Misinformation research would benefit from identifying solutions that conservatives and liberals agree on. Examples include giving people the option, like on social media platform Bluesky, to turn misinformation moderation on or off. But Trumps executive order seeks to ban that research. Thus, instead of providing protections, the order will likely weaken Americans defenses. H. Colleen Sinclair is an associate research professor of social psychology at Louisiana State University. This article is republished from The Conversation under a Creative Commons license. Read the original article.
Category:
E-Commerce
Rewards programs are getting revamps all over the place. Earlier this week, JPMorgan Chase announced major changes to its Chase Sapphire Reserve rewards credit card. That announcement came shortly after competitor American Express teased an upcoming overhaul to its premium Platinum rewards card. And now, one of Americas most popular cruise lines, Carnival, has announced it will radically shake up its rewards program come 2026. Heres what you need to know about Carnival Cruise Lines upcoming overhauled rewards program. Bye-bye, Very Important Fun Person (VIFP) Carnival Cruise Lines current rewards and loyalty program is called the Very Important Fun Person (VIFP) program. The program has been in existence since 2012 and is beloved by many Carnival Cruise enthusiasts. One of the reasons frequent Carnival customers love the VIFP program is its simplicity. You earn points based on how many days you cruisethats it. Theres no complicated earning structure or confusing rules. The total number of these points based on the total number of days cruised slots you into one of five groups: blue, red, gold, platinum, or diamond. The higher the group (diamond is tops), the more perks you get, including everything from free bottles of water to priority reservations at Specialty restaurants. However, Carnival has now announced that the VIFP program is being discontinued. Its last day of operation will be May 31, 2026. Hello, Carnival Rewards Carnivals VIFP program is being replaced with a new rewards and loyalty program called Carnival Rewards. The program will officially launch on June 1, 2026, and it represents a radical departure from the current VIFP program. One of the biggest changes to the new Carnival Rewards program is that points are no longer earned based on how many days you cruise with the line. Instead, points will be earned based on what you spend with Carnival Cruise Line. This spend includes the cost of tickets, drinks you buy on board, and even what you spend in the ships casino. Points can also be earned through the use of a new Carnival Rewards Mastercard, which will be debuting. However, you wont just be collecting points to move up the group rankings anymore. The new Carnival Rewards program will use a dual-earning structure. The dollars you spend will earn you both stars and Carnival Rewards points. The stars will dictate which group you are slotted intothere are only four this time: red, gold, platinum, or diamond (blue is going away). The more prominent the group, the better perks you get, such as embarkation and debarkation priority, if you are a member of the diamond group. The Carnival Rewards points you earn can be redeemed for anything from cruise fares, transfers, onboard purchases, and more. Why is Carnival revamping its rewards program? The business answer to this question is that Carnival Cruise Line likely believes that the revamped rewards program will lead to a better bottom line. As points and group tiers are now dependent on what you spend, customers who are keen to collect Canrival Rewards points may be more willing to spend more onboard to keep earning those points. But Carnival would likely argue that the revamped program is now a little fairer than its current VIFP program. Under the VIFP, Carnival customers would receive the same amount of rewards regardless of whether they paid for the lowest or highest class of cabin on the ship. Thats because the rewards were simply linked to the number of days cruised. The new program will mean that customers who book higher-priced cabins will now earn a higher number of reward points. But there is some bad news for customers under the new Carnival Rewards program, no matter how much they spend: the new status tiers they earnred, gold, platinum, and diamonddont last forever. Under the VIFP program, when a customer earned a status, they could only move up, not down. That means someone who earned diamond status would keep it for life. But the new status program will see customers lose their status after two years. And the points earned to achieve a status must be earned within a two-year period; otherwise, they do not count towards a status upgrade. In a FAQ about the new rewards program, Carnival says that this move is consistent with loyalty programs across the travel industry. The company says it recognizes that change can be difficult, but the current program based on cruise frequency makes it difficult to properly recognize our loyal guests. Carnival stock price still well below pre-pandemic levels The owner of Carnival Cruise Lines, Carnival Corporation & plc, filed its most recent earnings report on March 21 for the first quarter of 2025. The company announced that its Q1 revenues increased by over $400 million from the same quarter a year earlier, reaching $5.8 billion. At the time, Carnival Corporation & plcs CEO, Josh Weinstein, boasted that the quarter was truly characterized by outperformance. That growth is something investors are undoubtedly happy to see, especially after Carnival, like all other cruise lines, took a major hit in 2020 following the outbreak of the pandemic. In January 2020, Carnival Corporation & plcs stock (NYSE: CCL) was trading at above $50 per share. But by April, it had fallen to below $8 per share. As recently as October of 2022, CCL stock was trading below $7 per share. However, the stocks fortunes have improved as the cruise industry has slowly recovered from the pandemic’s impact, receding from peoples memories. As of yesterdays close, CCL stock is trading at above $23 per share. However, that is still down more than 5% since the beginning of the year. We wont know whether Carnivals new rewards program will have a material impact on the companys business, and thus its stock price, until it launches next year.
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||