 |
|
|
|||||
There are more possible NBA schedule combinations than there are atoms in the sun. That’s not hyperboleit’s the mathematical reality facing anyone trying to arrange 1,230 games across 30 teams over six months while satisfying TV networks, player safety rules, arena operators, and competitive fairness requirements all at once. This impossible puzzle is exactly what Fastbreak AI, a 30-person startup out of New York, has built its business around. Fastbreak’s AI software now powers scheduling for more than 50 professional leagues globally, quietly controlling when billions of dollars in sporting events hit your calendar. “I’m always amazed when we produce a playable schedule,” Fastbreak cofounder and CEO John Stewart says. “It’s a nearly impossible set of math problems. We’re considering billions upon billions of possibilities.” Map Anything Stewart’s path to sports scheduling began with a $250 million exit. His previous company, Map Anything, was acquired by Salesforce in 2019 for its field service optimization technology, which used the same mathematical principles that would later power Fastbreak. While still running Map Anything, in 2016, Stewart began recruiting two KPMG consultants, Chris Groer and Tim Carnes, who had built the NBA’s scheduling system, with the promise that hed eventually start a company dedicated to sports scheduling. When that occurred in June 2022, the timing was fortuitous. The NBA needed help scheduling its new in-season tournament, but the team they had worked with at KPMG was now at Fastbreak, making it easy for Stewart to onboard the league as one of the company’s first major clients. ‘If everyone hates you equally, you’ve done your job’ What the trio of founders discovered with the NBA schedule was a system of staggering complexity. The NBA has 30 teams, and each arena has different constraints. The San Antonio Spurs, for instance, are the fourth priority in their own building, meaning concerts can be prioritized over games. The Lakers’ venue hosts the Grammys and other marquee events each year, forcing the team to play on the road for certain stretches. Media partners pay billions for specific matchups to be in prime time and on marquee dates. Player safety rules prevent back-to-back games over 350 miles apart. And on top of that, each team is allowed to make requests. The Miami Heat, for instance, prefer to play at home during Art Basel. Still, not all requests can be granted. “It’s the art of managing disappointment equally,” Stewart says. “If everyone hates you equally, you’ve done your job right.” The challenge isn’t just mathematicalit’s diplomatic. Fastbreak’s platform gives different access to arena managers, media partners, and team executives, who can each enter requests into the system for consideration. When changes happen, the AI is designed to minimize collateral damage while accommodating whoever needs the adjustment. Schedule repairthe logistics game changer The traveling salesman problem is a classic mathematical and logistical challenge about finding the most efficient route through multiple cities. Computer scientists can solve that for hundreds of thousands of locations. But even a simplified sports scheduling problem featuring just 10 teams? “People have written many PhD dissertations on it and still not solved that problem to optimality,” Groer says. But even after clearing that mathematical hurdle, the job isn’t done. When the LA wildfires and Gulf Coast hurricanes disrupted games this year, forcing the NBA and NHL to reschedule, it triggered what experts call cascading optimization crises. A single venue change can force adjustments to hundreds of other games due to ripple effects across team travel schedules, TV contracts, and competitive balance requirements. Fastbreak’s schedule repair function suggests optimal fixes in minutes using what Stewart calls “warm starting”beginning from the current state rather than rebuilding from scratch. Think of it like GPS rerouting when there’s traffic, but infinitely more complex. When one game gets moved, the AI instantly recalculates how that affects every other game and the 500-plus other rules, then suggests the least disruptive solution to minimize collateral damage to uninvolved teams. The art behind the science Fastbreak’s breakthrough isn’t just computational power. It’s incorporating machine learningteaching AI systems to understand the subjective art of what makes a good schedule. League executives manually rate thousands of past road trips on a 110 scale, teaching the system what constitutes quality travel patterns. A trip hitting multiple East Coast cities in logical geographic order might rate an 8 or 9. A chaotic journey ping-ponging across time zones could get a 3. “You actually have to give business users a user experience where they can train this model and teach it the meaning of ‘good,'” Groer says. “You can never just provide all these trips to an AI model because it would immediately bias to ‘this trip’s been done in the past, therefore it must be acceptable.'” The result is that the AI has become more consistent than human experts were with each otherwhen multiple league officials rated the same trips, the AI’s ratings were closer to each expert’s opinion than the experts were to each other. Fastbreak’s AI uses a sophisticated scoring system to balance competing demands, weighing different violations based on league priorities. A hard constraintlike preventing teams from playing back-to-back games over 350 miles apartmight carry infinite penalty points, essentially making such schedules impossible to generate. Softer preferences, like avoiding Monday home games, carry smaller penalties that can be traded off against other benefits when the AI is trying to optimize the overall schedule. All told, professional schedules can have more than 500 different “rules,” each with carefully calibrated penalty weights to ensure accurate prioritization, and teams get point allocations for special requests. They might have 2,000 points to distribute across their wish list, creating a market-like system where they must prioritize what matters most. Beyond the big leagues Fastbreak now powers scheduling for the NFL, NBA, WNBA, NHL, and Major League Soccer, plus top college conferences like the SEC, ACC, and Big East. But Stewart sees an even bigger opportunity in youth sportsa $40 billion annual market where parents juggle multiple apps and constantly changing schedules. “I bet on your phone you’ve got nine different apps for those different sports, and I bet yu hate them all,” Stewart says, describing a frustratingly common experience for sports parents. In June, the company launched Fastbreak Compete, which has integrated the same AI scheduling engine it offers professional leagues. As of June, the software is used by 12 youth sports organizations, with commitments from over 40 more for 2026. Fastbreak’s strategy is to use its professional-grade technology as a hook, then expand into adjacent services. Fastbreak Compete creates schedules, but also serves as a one-stop shop for parents, as it consolidates scheduling, communications, travel booking, and payments into one platform, eliminating the app-juggling nightmare and providing real-time updates when tournaments inevitably change. Fairness, not perfection When the NBA season tips off, Stewart and his team will already be preparing to work on next year’s schedule, starting with arena availability collection in November. Its the first step toward the ultimate goal of building a schedule that is not just operationally efficient, but also fair. Fastbreak’s algorithms continuously monitor dozens of metrics: total travel miles, home weekend games, back-to-back frequency, rest advantages, and countless other factors that could create competitive imbalances. The extensive fairness metrics help ensure that when schedules are released, every team has roughly equal advantages and disadvantages across multiple dimensions. “If you’ve done your job right, everyone will find something to complain about, Stewart says. But the complaints will be equally distributed.” As leagues continue evolvingadding tournaments, managing global events, negotiating increasingly complex media dealsthe optimization challenges only intensify. For an industry built on competition, perhaps the ultimate victory happens behind the scenes, in algorithms that ensure the playing field remains level, one perfectly balanced constraint at a time.
Category:
E-Commerce
Apples iOS 26 for iPhone is now available to download. The star of the operating system is the all-new design language known as Liquid Glass, which introduces a translucent layer to various interface elements that mimics the physical properties of real glassincluding the way light shines and refracts through the material. But iOS 26 is about more than just a pretty new digital face. The new operating system also adds some compelling new productivity features that make Apples smartphone a more helpful tool for work than ever before. These are the top five new productivity features Apple has just added to the iPhone. 5. Only take the phone calls you want [Photo: Apple] One of the most common interruptions to our workflow is phone calls. And given that U.S. consumers received over 4.1 billion robocalls in August alone, the chances are good it might be spam. Yet this isn’t always the case. One way to distinguish between a call you need to take and one you don’t is by glancing at the iPhones caller ID. But this only works if you recognize the number calling. If the number is unknown, you need to pick up to see who is on the other lineand what theyre calling for. Now, with the new operating system, your iPhone gains a feature called Call Screening, which enables your iPhone to answer the unknown call for you by playing a message for the caller, which asks them to identify themselves and state a reason for their call. Once they do this, your iPhone will alert you to their response so that you can decide whether to take the call. If you decline, the unknown caller will be sent to voicemail. To enable Call Screening in iOS 26: Open the Settings app. Tap Apps. Tap Phone. Under the Screen Unknown Callers heading, choose the level of Call Screening you desire. The three levels are Never (disable Call Screening), Ask Reason for Calling (your iPhone will collect the caller’s name and their reason for the call before alerting you), or Silence (all calls from unknown numbers will go right to voicemail). 4. Create calendar events with a screenshot [Photo: Apple] For many conferences or retreats, event organizers will email poster invites for the event, which include a photo of the venue, date, time, and location. Or sometimes theyll post the invite to their social media feed or webpage. Prior to iOS 26, you had to enter this event information into your calendar manually. However, with the new OS, you can now take a screenshot of the invite, and your iPhone will extract the relevant information from it and add it to your calendar with just a few taps. This greatly simplifies the tedious process of manually adding invites to your calendar. Heres how: With the event invitation on your screen (whether it’s an image attached to an email, a webpage, or a social media post), take a screenshot on your iPhone by pressing the volume up and power buttons simultaneously. iOS 26 will detect the event information in the screenshot and display a Add to Calendar button below it. Tap this button. A pop-up will appear showing the event information that has been extracted from the screenshot. Tap the Create Event button to add the event to your calendar. You’ll now see a listing for the event in your Calendar app. 3. Make your iPhone wait on hold for you [Photo: Apple] With iOS 26, waiting on hold is a thing of the past, thanks to a new feature called Hold Assist. With Hold Assist, your iPhone can wait on hold for you while you continue with your work. When you finally get connected to someone, your iPhone will ring you again to let you know they are on the line. To use Hold Assist in iOS 26:/p> Open the Settings app. Tap Apps. Tap Phone. Ensure the Hold Assist Detection toggle is turned on (green). Now, dial a phone number as you normally would. When you are placed on hold, tap the Hold button on your iPhones dialer screen. Your iPhone will notify you when its time to pick up the phone again to talk to a real person. 2. Put AI to work as your foreign language interpreter [Photo: Apple] Business is global, and that means that we may often find ourselves needing to communicate with people who dont speak the same language as we do. Now, in iOS 26, your iPhone can act as your personal digital interpreter thanks to a new feature called Live Translation. The feature uses Apple’s AI platform, Apple Intelligence, and works across all major iPhone communication apps, including Messages, FaceTime, and Phone. In Messages, text conversations will be cross-translated in real-time. In FaceTime, real-time caption translations will appear on your screen. And in the Phone app, the iPhone will actually speak the translated conversation for you. Live Translation is enabled by default and will kick in whenever your iPhone detects you’re communicating with someone who is using a different language from yours. However, this feature only works on iPhones that support Apple Intelligence, which includes the iPhone 15 Pro and later models. 1. Customize folders for better file management [Photo: Apple] Apples Files app is a robust file management system on your iPhone, enabling you to store and access documents and other essential files for your work. Before iOS 26, the folders in which these documents were stored and sorted all looked the sameplain and pale blue. However, in iOS 26, Apple has added the ability to customize the look of each folder, allowing you to create custom designs that make it easy to pinpoint the folder containing the files you need at that moment. In iOS 26, you can customize the folders in the Files app in numerous ways, including by color and by adding a symbol or even an emoji to the front of the folder. Heres how: Open the Files app on your iPhone. Long-press on a folder you want to customize and tap Customize Folder & Tags . . . from the pop-up menu. Select the Tags button, then choose a color to give the folder a new hue, and tap the checkmark button. Now, choose a symbol you want to add to the front of the folder (Apple provides hundreds of options). Alternatively, tap the Emoji button and select an emoji you want to add to the front of the folder. Tap the blue checkmark button when done. Best of all, after customizing your folder to your liking, the new style will sync across all your devices using iCloud Drive, so itll look the same on your iPhone, iPad, and Mac. iOS 26 is now available as a free download. It requires an iPhone 11 or later to run.
Category:
E-Commerce
In a new report, OpenAI said it found that AI models lie, a behavior it calls scheming. The study performed with AI safety company Apollo Research tested frontier AI models. It found “problematic behaviors” in the AI models, which most commonly looked like the technology “pretending to have completed a task without actually doing so.” Unlike hallucinations, which are akin to AI taking a guess when it doesn’t know the correct answer, scheming is a deliberate attempt to deceive. Luckily, researchers found some hopeful results during testing. When the AI models were trained with “deliberate alignment,” defined as “teaching them to read and reason about a general anti-scheming spec before acting,” researchers noticed huge reductions in the scheming behavior. The method results in a “~30× reduction in covert actions across diverse tests,” the report said. The technique isn’t completely new. OpenAI has long been working on combating scheming; last year it introduced its strategy to do so in a report on deliberate alignment: “It is the first approach to directly teach a model the text of its safety specifications and train the model to deliberate over these specifications at inference time. This results in safer responses that are appropriately calibrated to a given context.” Despite those efforts, the latest report also found one alarming truth: When the technology knows it’s being tested, it gets better at pretending it’s not lying. Essentially, attempts to rid the technology of scheming can result in more covert (dangerous?), well, scheming. Researchers “expect that the potential for harming scheming will grow. Concluding that more research on the issue is crucial, the report said, “Our findings show that scheming is not merely a theoretical concernwe are seeing signs that this issue is beginning to emerge across all frontier models today.
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||