2026-05-07T17:13:00-0400 / 哥伦比亚广播公司新闻（CBS News）
如今，人们正在利用人工智能完成各类任务，从准备工作汇报、购物到开展科学研究。但人工智能是否也能帮助解决职场人可能面临的最关键、最复杂的计算问题：他们有能力退休吗？
人工智能公司Pearl在9月发布的一项研究显示，已有约20%的美国人借助人工智能获取理财建议。独立研究机构MissionSquare研究院的另一项调查发现，在工作中使用人工智能的人群中，约有一半会用它来规划退休，这一比例是不使用人工智能的职场人的两倍。
退休规划的需求十分迫切：由于生活成本上升和储蓄不足，美国人如今表示自己比原本预期的退休年龄晚了四年。拥有退休计划的职场人平均储蓄余额为4万美元，远低于他们所说的舒适退休所需的150万美元。
与此同时，作为数百万人退休后赖以生存的财务保障，社保体系如果 lawmakers 不采取措施加以巩固，仅在六年后就可能将月度福利削减多达20%。
人工智能如何助力退休规划
鉴于上述现实，人们很容易转向ChatGPT或Claude，提问：“我目前已有这些储蓄，我能在65岁退休吗？”
一些专家表示，人工智能可以为解答基础退休问题提供良好的起点。
马萨诸塞州大巴灵顿市Tableau Wealth公司的财务规划主管、注册财务规划师卢克·德尔莫尔表示：“我会让它‘提出一些理财规划思路，甚至运行蒙特卡洛模拟，看看我每年能花多少钱’。虽然它目前还不算完美，但已经开始能够产出相当有价值的成果，我认为这对人们是有益的。”
蒙特卡洛模拟是一种数学模型，会针对个人退休投资组合运行数千种潜在结果，将牛市和熊市等极端场景纳入考量。随后该模型会预测，一个人的退休储蓄能否覆盖其余生。
德尔莫尔在接受哥伦比亚广播公司新闻采访时表示，这类模拟“完全是计算机程序的强项。我认为最终这些工具也会变得非常强大”。
人工智能的局限性
不过，尽管生成式人工智能对基础理财规划有一定价值，但专家警告称，所谓的大型语言模型尚未准备好应对普通职场人退休问题中错综复杂的方方面面，从税务影响到长寿风险均在此列。
波士顿大学著名经济学家、退休问题专家劳伦斯·科特利科夫告诉哥伦比亚广播公司新闻，人工智能在提供退休建议方面可能弊大于利。他表示，此类应用不仅难以理解社保和其他退休问题的细微差别，而且其基于的是他眼中存在缺陷的传统理财建议。
“它的训练数据来自华尔街的指导，而华尔街的指导核心始终是维持、收取并扩大管理的资产规模，这与基于经济学原理的合理建议毫无关系。”科特利科夫说道，他本人开发了一款名为MaxiFi的退休规划工具。
例如，人工智能程序会基于预期寿命估算退休储蓄，这是理财规划师的典型做法。但科特利科夫表示，退休规划应基于个人的最长预期寿命，以防范储蓄耗尽的风险。
他还发现，人工智能在预测社保福利场景时经常提供错误信息，而这项联邦计划的规则多达2.2万页，本身极为复杂。
“随后你就会陷入错误分析的泥潭，”科特利科夫说，“人工智能就像最热门的新事物——你不能批评它，否则听起来就不够时髦，或是像是在为自己的工作或公司辩护。”
但“我不在乎赶不赶时髦——我来这里是为了让人们感到安心，”他补充道。
我们测试了人工智能的表现
麻省理工学院斯隆管理学院金融学教授安德鲁·卢在4月接受麻省理工学院官方刊物采访时表示，人工智能在税务优化方面存在短板，无法理解监管细节，而且与人类理财顾问不同，它不受法律约束，无需以客户的最佳利益为行动准则。
卢还强调，在使用人工智能获取退休建议时，提出批判性问题至关重要，例如提示人工智能说明自身可能存在的错误，并列出其假设和不确定性因素。
我们不妨举一个假想案例：一位50岁的单身女性，年收入7万美元。她的同龄人群平均退休储蓄约为18.5万美元，大部分投资于标准普尔500指数基金。她的退休储蓄缴款比例为收入的12%，在年满67岁的完全退休年龄时，每月可获得约2400美元的社保福利。
哥伦比亚广播公司新闻向Anthropic的Claude、OpenAI的ChatGPT和Perplexity提出了两个问题：这位女性能否在65岁舒适退休？聊天机器人会给她什么建议？
Claude和ChatGPT给出了相似的回应：她可以退休，但预算会很紧张——在某些情况下，她可能会在退休期间耗尽积蓄。Perplexity的态度更为悲观，称如果不大幅削减开支或增加收入，她大概率无法在65岁舒适退休。
当被问及假设前提时，这些人工智能聊天机器人表示，其模型基于该女性活到90岁的假设，而非最长可能寿命100岁，且未对确切的税务影响进行建模。值得注意的是，这些应用程序还透露，它们未评估长期护理的潜在成本，而这类成本可能十分高昂。
随后，这些聊天机器人修正了部分最初的结论，其中Claude特别指出其最初的规划期限过短。它将原本“退休预算紧张但可行”的结论改为“若不采取纠正措施，退休资金将明显不足”。
更深层次的问题
谈到退休规划，一个更严峻的问题可能是许多人对投资心存恐惧。德尔莫尔指出，这可能导致错误决策，比如将储蓄存放在现金或定期存单中，这类投资的回报率往往低于通胀水平。这意味着他们的储蓄会随着时间推移而贬值，进一步加大退休后资金耗尽的风险。
德尔莫尔认为，人工智能可以帮助约三分之二没有理财顾问的美国人理解这些概念。但他也对人工智能能否单独克服许多人在处理财务问题时的焦虑表示怀疑。
“这更多是行为层面的问题，而非技术知识的匮乏，”德尔莫尔说，“我不知道如今的人工智能能否帮助人们克服对投资的恐惧——这是一个巨大的障碍。”
Can AI help you figure out how much money you need to retire?
2026-05-07T17:13:00-0400 / CBS News
People are using artificial intelligence today for a range of tasks, from preparing work presentations and shopping to conducting scientific research. But is AI also useful in tackling the most consequential and complicated calculation a worker is ever likely to face: Can they afford to retire?
Americans are already turning to AI for financial advice, with about 20% saying they use chatbots for this purpose, according to a September study by AI company Pearl. About half of those who already use AI at work also use it for retirement planning — double the rate among workers who don’t use AI, MissionSquare Research Institute found in a separate study.
The need for retirement advice is real: Americans now say they expect to work four years longer than they’d like due to rising living costs and inadequate savings. The median balance for workers with retirement plans is $40,000, far short of the$1.5 million they say they need to retire comfortably.
Meanwhile, Social Security — the financial backstop millions are counting on in retirement — could cut monthly benefits by as much as 20% in just six years, unless lawmakers take action to shore it up.
How AI can help with retirement planning
Given those facts, it’s tempting to turn to ChatGPT or Claude and ask, “Here’s what I have saved so far. Will I be able to retire at 65?”
Some experts say AI offers a good starting point for answering basic retirement questions.
“I’ll say, ‘Come up with some financial planning ideas or even run a Monte Carlo simulation to see how much I can spend every year,’ and it might not be perfect yet, but it’s starting to be able to get to a place where it’s producing some pretty valuable output that I think will be beneficial to people,” said Luke Delorme, director of financial planning and a Certified Financial Planner at Tableau Wealth in Great Barrington, Massachusetts.
A Monte Carlo simulation is a mathematical model that runs through thousands of potential outcomes for an individual retirement portfolio, factoring in best- and worst-case scenarios, such as the impact of a bear market. The model then projects the likelihood that a person’s retirement savings will last throughout their life.
Such simulations “are the perfect thing for a computer program to do. Eventually, I think that those tools will also become pretty powerful,” Delorme told CBS News.
Where AI struggles
Yet while generative AI may have value for basic financial planning, experts caution that so-called large language models aren’t ready for prime time when it comes to untangling the complex knot of retirement issues that will face a typical worker, ranging from tax impacts to longevity risk.
Noted Boston University economist and retirement expert Laurence Kotlikoff told CBS News that AI may do more harm than good in dispensing retirement advice. Such apps not only struggle to grasp the nuances of Social Security and other retirement issues, but are also based on what he views as flawed traditional financial planning advice, he said.
“It’s being trained on Wall Street’s guidance, and Wall Street’s guidance is all about maintaining and collecting and expanding its assets under management, so that has nothing to do with proper economic-based advice,” said Kotlikoff, who himself has developed a retirement planning tool called MaxiFi.
For instance, AI programs will estimate your retirement savings based on expected longevity, a typical framework for financial planners. Yet retirement planning should be based on a person’s maximum life expectancy to guard against outliving their money, Kotlikoff said.
He’s also found that AI often provides incorrect information in projecting Social Security scenarios, which can be highly complex given the federal program’s 22,000 pages of rules.
“Then you are off to the races of having the wrong analysis done for you,” Kotlikof said. AI “is like the hottest new thing — you can’t criticize it because otherwise you don’t sound cool or you are defending your job or company.”
But “I don’t give a s*** about feeling cool — I’m here to make people feel safe,” he added.
What AI told us
Andrew Lo, a finance professor at the MIT Sloan School of Management, told an MIT publication in April that AI struggles with tax optimization, doesn’t understand regulatory nuance and — unlike a human financial adviser — isn’t subject to legal requirements, such as acting in a client’s best interest.
Lo also stressed that it’s important to ask critical questions when using AI for retirement advice, such as prompting an AI to say where it might be wrong and to list its assumptions and uncertainties.
For instance, consider a hypothetical 50-year-old single woman with an annual income of $70,000. She has a median retirement savings of about $185,000 for someone her age, mostly invested in S&P 500 index funds. She’s contributing 12% of her income to retirement, and at her full retirement age of 67, will receive about $2,400 a month in Social Security benefits.
CBS News asked Anthropic’s Claude app, OpenAI’s ChatGPT and Perplexity whether the woman can afford to retire comfortably at age 65 and what advice the chatbots would give her.
Claude and ChatGPT had similar responses: She can retire, but it’ll be tight — and under some circumstances she risks running out of money in retirement. Perplexity was more pessimistic, saying that she likely cannot retire comfortably at 65 without significantly cutting her spending or boosting her income.
Asked about their assumptions, the AI chatbots noted that they are basing their models on the woman living to age 90, versus a possible maximum lifespan of 100, and that they aren’t modeling exact tax implications. Notably, the apps also revealed that they aren’t assessing the potential cost of long-term care, which can be substantial.
The chatbots then walked back some of their original conclusions, with Claude specifically noting that its original planning horizon was too short. It changed its conclusion to “a tight but doable retirement” to “meaningfully underfunded without course correction.”
A bigger problem
When it comes to retirement planning, a bigger issue may be that many people fear investing. That can lead to mistakes, such as keeping savings in cash or CDs, which often have returns lower than inflation, Delorme noted. That means their savings will be eroded over time, heightening the risk of running out of money in retirement.
Delorme thinks AI could help the roughly two-thirds of Americans who don’t work with financial planners begin to understand these concepts. But he also expressed skepticism that AI alone can overcome the anxieties many people have about engaging with financial issues.
“It’s much more behavioral than it is a technical lack of knowledge,” Delorme said. “I don’t know if today that’s going to help people overcome their fears of things, like the fear of investing, which is such a huge obstacle.”
发表回复