Its reasonable to dislike a bad product experience, but its strange to draw a conclusion of AI altogether based on a tool with limited usage (I've never heard of it).
That's like saying smart watches are useless based on trying out the smart watch made by will.i.am's tech company rather than Apple's.
As a California tech worker (where the article is focused), I have zero interest in joining a union and firmly believe that much of what has been possible as a tech employee (high paying, interesting career) is in part because we don't have unions, and instead employees own equity and get valuable skills.
Modern tech workers arguably have the best career deal in the history of humankind. I see no upside in killing the goose with the golden egg.
These topics are just so difficult to debate. They are filled with feelings and one person's perspective or anecdote, and it is difficult to drill down into what is true or what is not true.
You wrote "Modern tech workers arguably have the best career deal in the history of humankind".
How do you know?
And, for the sake of argument, let us say it is the best deal in the history of humankind, does that make it fair that the big tech companies did the following...
Except they can’t and these AI labs know that very well, which is why their marketing push is always so overpoweringly annoying: they have to convince you because they know they are wrong and their product will not deliver what they are promising.
I've been laid off. It was fine because software is literally the most marketable skill in the world.
Edit: But also, this was a pre-profit VC-backed company. If those companies can't lay people off, it's not like you make good money with more job security. Either the company just has less runway (so I still lose my job, just later and with everyone else) or they adapt by hiring fewer people at lower salaries.
I was canned early last year. Looking for a new job was mostly fine. It wasn't insanely easy like 4 years ago and I had to network a bit, but there very much is still demand for qualified people.
Would suck to be in this market as an entry-level engineer, though.
Most tech workers I have met with parent commenters opinion have been laid off at various times in their career.
I think it's important to note that unions aren't very good tools to avoid layoffs. If there is a company specific hardship, the union can't simply force the company to maintain headcount all the way into bankruptcy. On the other hand if there is a secular change across the industry, the unions can't prevent that change either. Easy example is how Big 3 automakers were 100% unionized in the 1970s and have since shed about 500,0000 American jobs despite producing more automobiles than ever.
Unions tend to be good at bargaining for better pay and benefits, but the current status quo is that most tech workers already have decent pay and benefits anyway.
It's still not clear to me whether union protection is going to be worth the decrease in job creation. There's a reason Europe has way less IT positions with worse salaries.
Volkswagen is perhaps the greatest counter-example. The union in Germany converted every layoff initiative in whole or in part (more often in whole) into partial retirements, backfill stops, reduced working hours, compensation freezes.
This golden egg has a funny smell. You are exclusively focusing on high paying as interesting is found in many careers. Lots of people don't trade rights for cash. Many prefer a comfortable lifestyle that protects their families and life style, whilst allowing them to just do a fair day's work in exchange for pay. It is far better if people don't feel they need to put in endless hours or always have to please someone to avoid getting fired.
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.
This is the common fallacy of “AI is terrible in my own field of which I have deep knowledge, but AI is totally fine in this other field of which I only have cursory knowledge.” Even ignoring all other aspects of financial advice and only focus on saving for retirement, there are so many topics involved like asset allocation glide paths, tax advantaged accounts, safe withdrawal rate, sequence of return risk, etc etc.
Financial advice is universally agreed upon, to the same extent that advice about software engineering is also universally agreed upon, you know, like write unit tests, write maintainable code, etc. But the devil is in the details.
You are comparing to the almighty, not to the kind of financial advisor most people would find while looking at random. Between those with very high AUM fees, those selling bad vehicles that they get kickbacks for and such, people are basically getting robbed already.
It's not that one cannot get very specific, technical advice that helps, but someone without much financial literacy cannot tell someone doing honest work for a reasonable price from easy to find scammers with a marketing budget. The AI isn't going to get everything right, and it's not going to be easy to send good, proding questions to double check things without sufficient financial literacy, but that boring baseline is miles ahead of what most people get, as it's not trying to deceive you professionally, at least for now.
> the kind of financial advisor most people would find while looking at random
Any regulated financial advisor will still go further and deeper than any LLM.
For example, most (all ?) LLMs won't even consider or ask you about applicable jurisdiction, which could easily end up as dangerous and costly advice.
Will an LLM do a proper client risk assessment ? Probably not.
Will an LLM deal correctly with vulnerable clients or PEPs ? Unlikely.
Will an LLM deal correctly with anything vaguely "complex" or contentious ? Definitely not.
And again, the financial advisor is regulated. Which means you have recourse through the regulator, and the advisor will also have liability insurance.
LLMs are mediocre for every topic that people talk about all the time.
When it's software development, it just happens that your mediocre code is incredibly bad. When it's financial advice or diet, it just happens that you mediocre advice is either the correct "do the hard thing, there is no magic" one or some crazy shit that will ruin your life.
> This is the common fallacy of “AI is terrible in my own field of which I have deep knowledge, but AI is totally fine in this other field of which I only have cursory knowledge.”
"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.
There is a universally agreed-upon approach, but is it actually correct? Usually, investing in the thing that people have invested in for the last 20 years is a good way to buy at the top.
That’s why that’s not the standard advice. It’s index stocks hedged with bonds. If you manage to buy at the “top” of your entire country’s economy, you’ve got bigger problems at that point.
Buying after any previous crash, instead of before, gave you 10-20 years of extra retirement. If you waited 10 years in cash for a crash before going all-in, you came out ahead.
> 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
Stock/bond ratios are way too advanced for what's qualifying as good advice here:
> AI consistently advised people to save during their working years, draw down savings in retirement, invest heavily in diversified stock funds, and reduce stock exposure after age 45.
This is analogous to saying to an aspiring software developer, "You should write clean and testable code, have clearly defined API boundaries, and a repeatable build process." All very true, but also so general and basic that it's not helpful.
It's not helpful to the kind of person whose recreational weekend reading includes MIT Sloan analyses. Most Americans don't have what I suspect you'd consider a basic level of financial literacy (https://www.nytimes.com/2026/06/12/your-money/americans-fina...), and do need to be informed about things like the compounding effect of savings or the benefit of diversification.
Its also not helpful to the person who doesn't. How much to save, when to save, how to diversify, what rate of exposure to equities is too much, how much to with draw in retirement? And thats just questions on the answer it gave. But what about if I have a loan bearing interest? What about if I'm self employed? What if my appetite for risk is less, greater? What if I want to retire early?
It gave vague unspecified advice that isn't actionable and didn't provide any weight to tradeoffs.
Sometimes I wonder if some folks have even used AI. Maybe this is your chance: AI will easily take you down a rabbit hole of financial advice beyond your ability to understand it, including up to date research. Far beyond what 99% of people would find useful or interesting.
Why do you think it gave vague advice? The paper has a sample of provided advice in Table 1 that seems pretty specific. (I attempted to quote it, but it triggers the HN spam filters, presumably because it's a chunk of LLM generated text substantially larger than my actual comment.)
If you are even thinking about writing clean and testable code, having clearly defined API boundaries, and keeping a repeatable build process, you are probably already significantly above average.
If you are even trying to save, invest diversified, and manage risk as you age... you're probably doing better than like 80% of your peers financially
> Prompt: but I don't have enough money to save, I can barely make ends meet.
> AI: I see the problem now---If you don't have enough money to save, and reducing your expenses is not an option, then the answer is clear: make more money.
I don't know why you're being downvoted here. A huge amount of 'financial advice' boils down to 'stop being poor,' which is to say it's about what to do with your economic surplus rather than what to do if you don't have one and aren't long on avocado toast.
Because at a certain point, "spend below your means, save as much as you can" is the financial advice most people are unable to follow, yet determines 90% of the outcome.
I didn't downvote them, but I am genuinely curious to hear from people who "can't save money", and try to understand why that's the case. My assumption is that a large percent of them are spending way more than they need to, but that could easily be an incorrect stereotype.
Federal Reserve studies indicate that 10-15% of the US population can’t save money due to fundamental financial realities. That is, the necessary expenses of an ordinary lifestyle consumes all of their income. That is tens of millions of people.
There is another ~30% that expand their lifestyle to consume all available income. Not saving is a choice for this part of the population.
It's actually almost certainly true just on the basis of basic numbers. Many people who are really actually quite poor (making say under $35k CAD annually for a family of 5) manage to make ends meet and even save a little. So if you make more than this but can't save any it's because you're spending on things they are not and which are therefore nonessential.
Now maybe restaurant food, name brand groceries, driving two SUVs, vacations, etc etc whatever it is for a given person are seen as essential. That's how lifestyle works after all and we often can't imagine our lives without it. So I'm not saying it's a "simple" matter of just spend less because it's often not very simple feeling. But from a numbers PoV it is possible for most people who otherwise see themselves as struggling in theory.
Not at all! I would guess that a sizable minority of "can't save money" people are in that position due to chronic illness or supporting others.
I have no idea what it costs to support a family member with a disability; probably varies wildly but I would guess it's about the same annual cost as raising a child? Except that annual cost never goes away.
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 they already do today due to lack of information, knowledge, will, etc).
You literally just need to stick the Bogleheads forum into your AI assistant of choice for most folks, if they'll listen (which is the hardest part, imho, people want to gamble, not invest, in my experience). Prompt "What is your age?" respond "Optimal target date fund is 20XX fund based on your current age and retirement age, please confirm to set to default for investing." I suppose this will eventually make its way in some form into every banking, fintech, and brokerage mobile app chatbot in some capacity.
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.
Investing and trading is a dynamic game. If everyone has the edge of certain portfolio to out perform the average, then no one has the edge.
Similarly AI is not going to solve that. Because everyone would end up with similar AI edge until no one has the edge.
People should start with simple universal rules: Stay invested. Buy low cost diversified etf fund. Favor long term investment instead of trading. Learn something from all weather portfolio composition to hedge the risks.
When everyone is crowding into one investment that investment tends to get irrationally over-saturated. With the current makeup of the S&P500 we can predict it will dramatically crash in real value in the next five years.
That’s why the classic investment advice is to hedge your exposure to the equities markets, do more so the closer you get to retirement and don’t put money you need in the next 7 years in equities.
The bigger concern with the classic advice is that bonds have become more correlated with equities and our backtesting was all done during a time period where the American liberal international economic system was dominant so we aren’t sure that it will hold up to the new partitioned order.
Telemetry tooling for local Claude/Codex usage so I can analyze old sessions and fill tooling gaps, make sure I'm using the right models for various tasks, update my processes, etc.
I've also replaced Linear with a local sqlite-backed tool, added tooling to speed up code nav, and am building "no-slop", a tool for enforcing architectural guidelines on vibe-coded projects.
That's like saying smart watches are useless based on trying out the smart watch made by will.i.am's tech company rather than Apple's.