Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AI financial advice is surprisingly good, especially if you ask right questions

mitsloan.mit.edu

Post reply on HN
Print view View original

371–380 of 443 posts

Re: AI financial advice is surprisingly good, especially if you ask right questions

#371

Post by brunoarueira »

MikeNotThePope wrote: post #52
tokioyoyo wrote: post #45

Earlier quoted context omitted.

Models aren’t trained as much from random internet text as they used to in pre-2024 era. Like they are, but specialized datasets get more attention.

Your model/harness will indiscriminately do web searches to get answers. I believe that’s where the real risk is.

I don't know for sure, but will it require something like with high "SEO" rankings to fake the reputation of the sources? So if the LLM search for a specific spreaded knowledge splitted across many bad sites, it can poison the model maybe.

Re: AI financial advice is surprisingly good, especially if you ask right questions

#372

Post by groundzeros2015 »

inigyou wrote: post #362
groundzeros2015 wrote: post #356

Earlier quoted context omitted.

Many people go through a “homeless” period of not having a permanent address. Couch surfing. Motels etc. It’s not fun. It’s very stressful. Our systems have a positive feedback loop against financial instability. However there is another category of homeless which is a person pushing a shopping cart on the side of the road. And this is an exteme level of dysfunction and despair. I saw a tweet that summarized this iss…

If you took an average person and then forced them to stop paying rent or anything more expensive than their current rent (so no motels)—the scenario being discussed here—which category of homeless would they likely end up in?

What are you talking about? What is “rent”? You can rent an individual room. You can rent a double-wide trailer. You can move to Kentucky. You can work 40 hours (guaranteed 40-50k year).

> stop responding to all incentives and deliberately go against the grain of every system and form of aid.

Yep, that’s a pretty unique kind of person.

Re: AI financial advice is surprisingly good, especially if you ask right questions

#373

Post by Kirby64 »

francisofascii wrote: post #360
DonsDiscountGas wrote: post #65

Earlier quoted context omitted.

Lol. You don't even need AI for that 99% boiler plate. Save 6-12 months of expenses in cash, DCA the rest into total market stock index funds. But people still pay expensive advisors to get worse results.

The tricky part is which total market index funds? The S&P based ones are too AI focused, and don't give you the diversification they once did. You also don't want to invest in just one countries stock market. And bonds, should be a mix of maturity, governments (not just the US), corporate, etc.

Total market is total market. Investing in a “total market minus X” fund means you lose exposure to X (whatever that is) for good or bad. That’s your judgment call, but not really an edict. Same problem with “you can’t just invest with US stocks” - US stocks are global at this point, and have significant exposure to the international market. By investing in specific international funds you’re more or less saying “international minus the US”.

You can change your weighting if you want to, that’s your prerogative, but don’t be surprised if it doesn’t lead to nearly as good of average returns.

Re: AI financial advice is surprisingly good, especially if you ask right questions

#374

Post by MBCook »

wingworks wrote: post #222
MBCook wrote: post #182

Earlier quoted context omitted.

Yeah just off the top of my head I’d expect Discover and AmEx to be around 3.5%. Apple is at 3.4%. I moved away from near 0% savings accounts more than 20 years ago, it’s amazing to me it’s still so common. You don’t have to try very hard or go wrong to someone you’ve never heard of to get a good rate.

The rate goes up and down with inflation, high inflation, high interest - low inflation, low interest. But even then, yes some banks still offer no or 0.5% accounts.. because they can, and many people can't be bothered to figure out a better option, or "trust" there bank and don't want to move. (or the bank has high interest account, but make it complicated to use)

When I first got a real high rate savings account it was 5%. I think my normal bank paid an incredible 0.05% at the time.

You’re right it varies and that has been obvious lately as I seem to be getting monthly email warning me my rate would be going down.

Re: AI financial advice is surprisingly good, especially if you ask right questions

#375

Post by throw0101a »

tdrz wrote: post #350
throw0101a wrote: post #329

Earlier quoted context omitted.

> I wouldn't recommend this after seeing how SpaceX was literally shoved down lots of people's throats. If you're going to buy a "total market" fund, then SpaceX is part of the market. There were strange financial things with GE, Enron, etc , and they were part of index(es): you have to take the good with the bad when it comes to human (economic) behaviour. Most stocks suck: * https://papers.ssrn.com/sol3/papers.cfm?…

To me, investing is NOT just getting the best outcome possible. I simply don't agree with some companies practices and therefore I don't want to invest in them. I believe I am better off if I live in a better society overall than if I have more money in a worse overall society.

> I simply don't agree with some companies practices and therefore I don't want to invest in them.

Understandable, but unless you buy the stock from them at IPO, you're not giving them money. I agree with Cullen Roche's four points on ESG investing; second one:

> 2) The secondary market is a bad place to enact change. The intelligent defense of ESG is “by reducing the demand for a stock we can increase its cost of capital and impact its operating performance.” This is true to some degree, but I think this is dramatically overstated. For instance, the firms in the S&P 500 are all large established firms that have more than enough capital to finance their operations. They aren’t using the secondary equity markets to fund their operations. In fact, most firms have so much capital that they’ve been net buyers of stock in the last 50 years. So, this puts the cart before the horse. The better way to think of public companies is to think of them like horse betting. We can bet on the horses, but secondary market purchases are just private exchanges, not cash issuance to firms. As a result, betting on the horses doesn’t change the outcome of the race. Similarly, our secondary market purchases and sales have a far smaller impact on the firm’s operations than we might think.¹

* https://www.pragcap.com/my-view-on-esg-investing/

Roche's point in 'doing good' with investing is to make as much money as you can and then fund the movements and organizations that you wish to succeed (worked for the Koch brothers and others of their ilk).

Re: AI financial advice is surprisingly good, especially if you ask right questions

#376

Post by foxtrot8672 »

toomuchtodo wrote: post #26
aprilthird2021 wrote: post #24

Earlier quoted context omitted.

> By comparison, financial advice is pretty simple, and there is a universally agreed-upon approach that most people should follow to maximize long-term financial health. What will AI do when those rules, which it's trained on their repetition so much, don't apply anymore? ~8% annual stock gains for the next 40 years may not hold and an 80/20 stock/bond ratio may not be as wise in upcoming decades

Models can be updated when foundations domain knowledge graphs are built on change. As of this comment, target date funds and pensions containing trillions of dollars adhere to the assumptions you mention (asset class allocation, growth rate and return assumptions, safe withdrawal rates ["Trinity study" aka ~4%/year], etc), and so consumers of AI provided guidance assuming these foundations could do much worse (as th…

Context mangement is necessary to fully realize the power of AI financial advice. Memories, financial history, and decisions need to made with the AI. This translates into the epistemic record of every answer and its how we do it. This allows us to precisely repeat AI experiments with a given set of inputs and is necessary to ensure a high quality output.

The paper talked about how an AI informed the usr to build a financial emergency fund. but, when the user lost their job, the AI completely forget it existed. This proves our theory that context management is the key to unlocking the full potential of AI financial advice.

Re: AI financial advice is surprisingly good, especially if you ask right questions

#377

Post by MBCook »

ozim wrote: post #328
eks391 wrote: post #138

Earlier quoted context omitted.

3% is becoming more common as of the last few years, at least in the US. I know several banks off the top of my head that offer 3.5% or higher (and more if you are a new customer) for their savings accounts. I would persuade people who use banks that haven't moved on from near-zero APY to move on themselves.

Most annoying on Reddit are people who write about "high yield savings accounts", those might have been there in 80's or 90's but I see right away those people are just LARPINg. 3% is nothing there are no "high yield savings accounts".

It’s still better than basically 0%.

Yeah they’re not going make you rich. It doesn’t take the place of investing. But its still better than a checking account for money you need to keep liquid.

Re: AI financial advice is surprisingly good, especially if you ask right questions

#378

Post by foxtrot8672 »

jamestimmins wrote: post #19

AI seems to struggle most when it has to make decisions with lots of trade-offs, especially where the context or implications of various decisions are nested, which is presumably why it struggles to write full software systems that are well-designed. By comparison, financial advice is pretty simple, and there is a universally agreed-upon approach that most people should follow to maximize long-term financial health.

"By comparison, financial advice is pretty simple, and there is a universally agreed-upon approach that most people should follow to maximize long-term financial health."

yes, but...although the fundamentals are basically the same that doesn't mean it translates into an actual plan for a user. you're still leaving the hard part up to the user instead of helping them form an actual plan and stick to it.

Re: AI financial advice is surprisingly good, especially if you ask right questions

#379

Post by inigyou »

groundzeros2015 wrote: post #372
inigyou wrote: post #362

Earlier quoted context omitted.

If you took an average person and then forced them to stop paying rent or anything more expensive than their current rent (so no motels)—the scenario being discussed here—which category of homeless would they likely end up in?

What are you talking about? What is “rent”? You can rent an individual room. You can rent a double-wide trailer. You can move to Kentucky. You can work 40 hours (guaranteed 40-50k year). > stop responding to all incentives and deliberately go against the grain of every system and form of aid. Yep, that’s a pretty unique kind of person.

I'm talking about this scenario we are discussing, where someone saves money by not paying rent. Which kind of homeless will they be?

Re: AI financial advice is surprisingly good, especially if you ask right questions

#380

Post by foxtrot8672 »

Financial advice, at least to us at Pendragon, is not simply "here are the rules of thumb for maintaining a healthy financial life" those are helpful guidelines, but not a plan. AI is helping us help users by building and implementing actual plans tailored to their specific use cases and situation.

Its the difference between "You spend too much on dining, you should be putting that money into a HYSA instead" vs "You spent $150 on a dinner this weekend to celebrate landing that new deal. It's slightly over-budget, but you're still well on track with the goals and plans we set up a week ago. No adjustments are needed."

Post reply on HN