Latest news with #Zimbardo
Irish Independent
4 days ago
- Business
- Irish Independent
Meta enters 20-year contract for nuclear energy as AI power demands increase
The parent company of Facebook, Instagram and WhatsApp will probably be paying about $80 (€70) per megawatt hour for energy from the Clinton plant in Illinois, according to Paul Zimbardo, an analyst at Jefferies, who made the forecast based on company guidance. That compares with the firm's estimate of at least $110 for Microsoft's deal for power from Pennsylvania's Three Mile Island plant. Constellation Energy operates both nuclear sites and hasn't disclosed details for either agreement, with chief executive officer Joe Dominguez saying in an interview that the contracts reflect 'the fair price of reliable clean energy', without providing more details. The pricing outlook for Meta and Microsoft is different because generating energy at Three Mile Island is requiring the restart of a reactor, Mr Zimbardo said. The cost to revive the plant is estimated at $1.6bn after it was shut in 2019 for economic reasons. Meanwhile, the Clinton site is already in operation. 'There's a difference between a restart and a normal, operating plant,' Mr Zimbardo said. Technology companies have shown that they're willing to pay hefty premiums for nuclear energy, which is generally more expensive than fossil-fuel generation. They're betting that the carbon-free electricity that comes from atomic plants will help them meet their ambitious climate goals even as demand climbs to power the AI boom. Wind and solar aren't available all the time, but data centres are humming around the clock, and 24/7 nuclear energy meets both power and climate requirements. 'We have to recognise that some things are going to cost more than others,' said Michael Polsky, CEO of the independent power developer Invenergy. He spoke at a national security conference in Washington on Tuesday. 'Nuclear will cost more than gas. So if we want nuclear power in the future, we have to admit it will cost more,' Mr Polsky said. ADVERTISEMENT Under the Meta deal, Constellation will invest in boosting Clinton's output. The company is also considering plans to build another reactor at Clinton, which already has federal approval for a second unit. Overall electricity prices in the region served by the Clinton plant have been climbing. Wholesale power climbed to $666.50 a megawatt-day for the grid operated by the Midcontinent Independent System Operator in an April capacity auction. That's up sharply from $30 a year earlier. The contract is Meta's largest power deal to date, according to Urvi Parekh, the company's global head of energy. The company is increasingly interested in nuclear to run its operations, and in December announced it was seeking proposals for as much as 4 gigawatts of new US reactors. So far it has received about 50 proposals from a range of companies, including Constellation. That initiative is aimed as bringing power onto the grid in the early 2030s, while the Clinton deal is seen as a near-term effort. With the subsidies set to expire, Meta and Constellation said they wanted to ensure the plant remained competitive.
Forbes
16-04-2025
- Health
- Forbes
How AI Can Reshape Access To Specialty Medications
Sally Ragab, CEO @ Neunetix. In 1969, a psychologist named Philip Zimbardo conducted a now-famous experiment. He parked two identical cars in two neighborhoods: one in the Bronx, a high-crime area, and the other in Palo Alto, a quiet, affluent community. The car in the Bronx was quickly vandalized, stripped within hours. The Palo Alto car sat untouched. But then Zimbardo broke one of its windows. Within days, the once-pristine car was equally wrecked. Zimbardo's "broken window theory" suggested that environments left unchecked can lead to systemic dysfunction. I see healthcare facing its own version of this phenomenon. Prior authorizations, which I see as an outdated and fragmented process, have become the "broken window" of specialty medications—neglected, frustrating and quietly draining resources. But unlike in the 1960s, we now have tools that can help. I believe AI offers a path to not just patch the system but to reimagine it entirely. Prior authorizations for specialty medications are often managed manually—often involving faxes and phone calls. Providers report spending 12 to 15 hours a week on these tasks, which can result in delayed treatment, lost revenue and burned-out staff. For a patient waiting on chemotherapy or a biologic for Crohn's disease, that delay isn't just inconvenient—it can be life-altering. AI isn't just a buzzword in this space; it's a practical tool that's already making an impact. Here's what I witness it enabling within the healthcare space: • Real-time eligibility checks by analyzing patient insurance details the moment a prescription is entered. • Predictive denial prevention, flagging incomplete or noncompliant requests before they're submitted. • Natural language processing to extract clinical data from electronic health record notes, lab results and attachments without manual entry. • Automated submission and tracking across payer portals, with proactive alerts on missing information. In my direct experience, I've seen AI reduce approval times from 10 to 14 days to as little as two to three days. That's not theory—that's from real clinics we've partnered with. But AI doesn't work in isolation. To be successful, providers should: 1. Start with high-friction areas (e.g., oncology, rheumatology and rare disease) where prior authorization is frequent and urgent. 2. Involve clinical staff early to identify workflow pain points. 3. Choose AI tools that integrate directly into your electronic health records and don't force a new UI. 4. Review audit logs regularly and tune your AI models to local payer patterns. AI in healthcare raises valid questions: Will it follow HIPAA? Can we trust it to make clinical inferences? The answer is yes, but only with oversight. The best AI tools are transparent, explainable and built with guardrails. Vendors should offer encryption, role-based access control and Systems and Organization Controls 2 (SOC 2) compliance. And just as important: Staff need training not just on how to use the tools, but on how to challenge and verify their decisions. Building off the need for your staff to challenge AI decisions, AI isn't here to replace people; it's here to take the weight off their shoulders. It's the assistant that never sleeps, never gets overwhelmed and doesn't forget payer rules. In the context of healthcare, I see it as a force multiplier enhancing the role of pharmacists, providers and care coordinators. When we introduced AI at a multisite provider group, the most surprising feedback wasn't just about time saved; it was how morale improved. Nurses and coordinators said they finally felt like they could focus on patients again. The future of prior authorizations won't be about eliminating them; it will be about making them invisible. The AI systems will know what's needed before we do. They'll draft documentation, catch gaps and smooth the back-and-forth. But it only happens if we adopt early and thoughtfully. Prior authorizations have been a source of friction for too long. AI gives us the power to fix the "broken windows"—not with duct tape but with real structural change. The next step is ours to take. Forbes Business Council is the foremost growth and networking organization for business owners and leaders. Do I qualify?
Yahoo
28-02-2025
- Yahoo
Review: Rethinking the Stanford Prison Experiment (opinion)
If you took psychology courses in college, you probably remember the "Stanford prison experiment," which monitored the behavior of 18 students assigned to play the roles of guards or inmates in a pretend penitentiary for a week. Stanford University psychologist Philip Zimbardo's 1971 study supposedly demonstrated that decent, ordinary individuals are apt to cruelly mistreat people when given authority over them, even when that authority is imaginary. The Stanford Prison Experiment: Unlocking the Truth, a three-part National Geographic documentary, casts doubt on that gloss by interviewing academics and former subjects who say Zimbardo misrepresented his methodology and the implications of his results. Although Zimbardo claimed he gave very little direction to the "guards," for example, recordings show that he and his colleagues encouraged harsh treatment of the "prisoners." The documentary suggests the so-called experiment is better understood as an improv game in which the subjects acted in ways they thought Zimbardo expected. Yet after torture at the Abu Ghraib military prison in Iraq came to light decades later, Zimbardo, a relentless self-promoter, glibly argued in media appearances and in his book The Lucifer Effect that his research anticipated such abuses by showing "how good people turn evil." Director Juliette Eisner gives Zimbardo (who died last October, after the documentary was completed) ample opportunity for rebuttal. To some extent, his answers reinforce the impression of slipperiness. But his defenses also highlight an ambiguity at the heart of his experiment: Were the subjects just play-acting, as some now claim, or is that description a self-deceiving rationalization for mortifying behavior? In the end, Zimbardo loses patience with Eisner's inquiry, declining to answer follow-up questions on the grounds that he has already thoroughly refuted his critics' arguments. Viewers can judge for themselves whether that is true. The post Review: Rethinking the Stanford Prison Experiment appeared first on
