> Should I replace sodium chloride with sodium bromide?
>> No. Sodium chloride (NaCl) and sodium bromide (NaBr) have different chemical and physiological properties... If your context is culinary or nutritional, do not substitute. If it is industrial or lab-based, match the compound to the intended reaction chemistry. What’s your use case?
Seems pretty solid and clear. I don't doubt that the user managed to confuse himself, but that's kind of silly to hold against ChatGPT. If I ask "how do I safely use coffee," the LLM responds reasonably, and the user interprets the response as saying it's safe to use freshly made hot coffee to give themself an enema, is that really something to hold against the LLM? Do we really want a world where, in response to any query, the LLM creates a long list of every conceivable thing not to do to avoid any legal liability?
There's also the question of base rates: how often do patients dangerously misinterpret human doctors' advice? Because they certainly do sometimes. Is that a fatal flaw in human doctors?
kevinventullo · 1h ago
Is anyone else getting tired of these articles?
“Area man who had poor judgement ten years ago now has both poor judgement and access to chatbots”
A4ET8a8uTh0_v2 · 36m ago
What it does remind me of is the story of a person, who were following GPS instructions religiously[1]. The clickbait is a thing, but part of it may be some level of societal concern that a good chunk of society will listen if you tell it what to do.
Part of me rationalizes it as 'not exactly a discovery', which on its own was not a big issue before we were as connected as we, apparently, are ( even if I would argue that the connection is very ephemeral in nature ). I am still personally working through it, but at which point is the individual actually responsible?
I am not saying this lightly. I am not super pro-corporate, but the other end of this rope is not exactly fun times either. Where is the balance?
So tired. None of these are newsworthy. There were plenty of people making stupid decisions (wether about their health or anything else in their lives) before AI existed, and there will be plenty of people making stupid decisions while AI exists as well.
MangoToupe · 57m ago
It would be a little less tiring if we were to prosecute the folks responsible.
RiverCrochet · 43m ago
It sounds like the rare condition is already punishing him, but maybe if it were known there would be additional legal penalities for misusing tech, it might result in less misuse. If it works we could do similar things with cars and weapons.
flufluflufluffy · 40m ago
The only one responsible is the man himself. Make a stupid decision, reap a stupid reward.
pryelluw · 43m ago
Curious how that would work. There’s definitely known and understood laws around malpractice, fraud, and incompetence. Are there laws that would allow a less-than-gifted individual, such as the one in the article, to sue for misinformation?
slicktux · 1h ago
Does that title seem like a cluster to anyone else? I tried rewording it in my head but only came up with a slightly better solution:
“Man develops rare condition after cutting consumption of salt do to a ChatGPT query”.
jleyank · 1h ago
How about “man gets bromine poisoning after taking ChatGPT medical advice”?
brokencode · 1h ago
We don’t know whether ChatGPT gave medical advice. Only that it suggested using sodium bromide instead of sodium chloride. For what purpose or in what context, we don’t know. It may even have recommended against using it and the man misunderstood.
topaz0 · 56m ago
Chatgpt doesn't give advice at all, but could say "after interpreting chatgpt output as medical advice"
neom · 1h ago
It's clunky but I understood it immediately, I presumed from the title that it was going to be about your title, however I also thought it was a bit clunky.
dfee · 1h ago
still haven't clicked in, but was confused.
still am, unless i re-interpret "do" as "due".
genter · 1h ago
I read it as "race condition". I was then trying to figure out what salt has to do with a race condition.
throwmeaway222 · 1h ago
You're absolutely right! If you stop eating salt you will become god!
neom · 1h ago
You're absolutely right! I was locked in with GPT5 last night and I actually discovered that salt is a geometric fractal containing a key insight that can be used by physicists everywhere to solve math. Don't worry, I've emailed everyone I can find about it.
topaz0 · 53m ago
I hope you used the llm to write the emails
foobarian · 1h ago
Why do you say that?
Hikikomori · 53m ago
You're absolutely right!
pryelluw · 56m ago
These AIs are taking away the jobs of psychics and other snake oil peddlers. How will the median person get their confirmation bias serviced when the AI becomes too expensive?
RiverCrochet · 41m ago
I remember the psychic infomercial craze of the early to mid-90's. Think Dionne Warwick's Psychic Friend's Network or the later Miss Cleo "Call Me Now" that aired on informercial spots everywhere in the 90's.
Your comment gave me a nightmare of that returning, but in AI form somehow.
pryelluw · 19m ago
They inspired my comment. I think it’s already happening, though. Even the coding agents operate a bit like psychics.
MarkusQ · 1h ago
LLMs don't think. At all. They do next token prediction.
If they are conditioned on a large data set that includes lots of examples of the result of people thinking, what they produce will look sort of like the results of thinking, but then if they were conditioned on a large data set of people repeating the same seven knock knock jokes over and over and over in some complex pattern (e.g. every third time, in French), what they produced will look like that, and nothing like thinking.
Failing to recognize this is going to get someone killed, if it hasn't already.
uh_uh · 1h ago
> LLMs don't think. At all.
How can you so confidently proclaim that? Hinton and Ilya Sutskever certainly seem to think that LLMs do think. I'm not saying that you should accept what they say blindly due to their authority in the field, but their opinions should give your confidence some pause at least.
dgfitz · 56m ago
>> LLMs don't think. At all.
>How can you so confidently proclaim that?
Do you know why they're called 'models' by chance?
They're statistical, weighted models. They use statistical weights to predict the next token.
They don't think. They don't reason. Math, weights, and turtles all the way down. Calling anything an LLM does "thinking" or "reasoning" is incorrect. Calling any of this "AI" is even worse.
hodgehog11 · 33m ago
If you have an extremely simple theory that debunks the status quo, it is safer to assume there is something wrong with your theory, than to assume you are on to something that no one else figured out.
You are implicitly assuming that no statistical model acting on next-token prediction can, conditional on context, replicate all of the outputs that a human would give. This is a provably false claim, mathematically speaking, as human output under these assumptions would satisfy the conditions of Kolmogorov existence.
dgfitz · 22m ago
Sure.
However, the status quo is that "AI" doesn't exist, computers only ever do exactly what they are programmed to do, and "thinking/reasoning" wasn't on the table.
I am not the one that needs to disprove the status quo.
phantom784 · 48m ago
But is the connection of neurons in our brains any more than a statistical model implemented with cells rather than silicon?
scarmig · 17m ago
You're forgetting the power of the divine ineffable human soul, which turns fatty bags of electrolytes from statistical predictors into the holy spirit.
fl7305 · 45m ago
An LLM is very much like a CPU. It takes inputs and performs processing on them based on its working memory and previous inputs and outputs, and then produces a new output and updates its working memory. It then loops back to do the same thing again and produce more outputs.
Sure, they were evolved using criteria based on next token prediction. But you were also evolved, only using critera for higher reproduction.
So are you really thinking, or just trying to reproduce?
hcdx6 · 1h ago
Are you thinking over every character you type? You are conditioned too by all the info flowing into your head from birth. Does that gauruntee everything your brain says and does is perfect?
People believed in non existent WMDs and tens of thousands got killed. After that what happened ? Chimps with 3 inch brains feel super confident to run orgs and make decisions that effect entire populations and are never held accountable. Ask Snowden what happened after he recognized that.
hodgehog11 · 43m ago
I hate to be that guy, but this is (a) little to do with the actual problem at hand in the article, and (b) a dramatic oversimplification of the real challenges with LLMs.
> LLMs don't think. At all. They do next token prediction.
This is very often repeated by non-experts as a way to dismiss the capabilities of LLMs as some kind of a mirage. It would be so convenient if it were true. You have to define what 'think' means; once you do, you will find it more difficult to make such a statement. If you consider 'think' to be developing an internal representation of the query, drawing connections to other related concepts, and then checking your own answer, then there is significant empirical evidence to support high-performing LLMs do the first two, and one can make a good argument that test-time inference does a half-adequate, albeit inefficient, version of the latter. Whether LLMs will achieve human-level efficiency with these three things is another question entirely.
> If they are conditioned on a large data set of people repeating the same seven knock knock jokes over and over and over in some complex pattern (e.g. every third time, in French), what they produced will look like that, and nothing like thinking.
Absolutely, but this has little to do with your claim. If you narrow the data distribution, the model cannot develop appropriate language embeddings to do much of anything. You could even prove this mathematically with high probability statements.
> Failing to recognize this is going to get someone killed, if it hasn't already.
The real problem as in the article is that the LLM failed to intuit context, or to ask a followup. While a doctor would never have made this mistake, the doctor would know the relevant context since the patient came to see them in the first place. If you had a generic knowledgeable human acting as a resource bank that was asked the same question AND requested to provide nothing irrelevant, I can see a similar response being made. To me, the bigger issue is that there are consequences to easy access to esoteric information for the general public, and this would be reflected more in how we perform reinforcement learning to assert LLM behavior.
henearkr · 47m ago
A weights tensor is very similar to a truth table or a LUT in a FPGA, it's just a generalization of it with real numbers instead of booleans.
And then again, would you say that you cannot build a (presumably extremely complex) machine that thinks?
Do you think our brains are not complex biological machines?
Where I agree is that LLMs are absolutely not the endgame. They are super-human litterary prodiges. That's it. Litterary specialists, like poets, writers, scenarists, transcriptors, and so on. We should not ask them anything else.
nimbius · 1h ago
yeah sure, but, did it enrich the shareholders?
cortic · 1h ago
I'm not sure humans are any different;
Humans don't think. At all. They do next token prediction.
If they are [raised in environments] that includes lots of examples of the result of people thinking, what they produce will look sort of like the results of people thinking, but then if they were [raised in an environment] of people repeating the same seven knock knock jokes over and over and over in some complex pattern (e.g. every third time, in French), what they produced will look like that, and nothing like thinking.
I believe this can be observed in examples of feral children and accidental social isolation in childhood. It also explains the slow start but nearly exponential growth of knowledge within the history of human civilization.
ofjcihen · 44m ago
That’s…completely incorrect.
I’m not going to hash out childhood development here because I’m not paid to post but if anyone read the above and was even slightly convinced I implore you to go read up on even the basics of early childhood development.
cortic · 12m ago
> I implore you to go read up on even the basics of early childhood development.
That's kind of like taking driving lessons in order to fix an engine. 'Early childhood development' is an emergent property of what could be cumulatively called a data set (everything the child has been exposed to).
MangoToupe · 58m ago
Sure, but you can hold humans liable for their advice. Somehow I doubt this will be allowed to happen with chatbots.
No comments yet
gdbsjjdn · 54m ago
The industry is really trying to make "the computer cannot be held responsible" a feature instead of a bug.
Sure the machine very confidently lies about dangerous things, but you shouldn't trust it. But you should employ it to replace humans.
The US medical system practically requires patients to steer their care among specialists. If the gentleman steered himself to a liver doctor, he’d hear liver advice. Psychologist, he’d talk about his feelings. Can one really blame him for taking it one step further and investigating whatever he was worried about on his own too?
Plus, if you don’t have some completely obvious dread disease, doctors will essentially gaslight you.
These researchers get up on a pedestal, snicker at creative self-help, and ignore the systemic dysfunction that led to it.
some_random · 1h ago
Is it just me or is the title kinda unclear?
>The patient told doctors that after reading about the negative effects of sodium chloride, or table salt, he consulted ChatGPT about eliminating chloride from his diet and started taking sodium bromide over a three-month period. This was despite reading that “chloride can be swapped with bromide, though likely for other purposes, such as cleaning”. Sodium bromide was used as a sedative in the early 20th century.
In any case, I feel like I really need to see the actual conversation itself to judge how badly chatgpt messed up, if there's no extra context assuming that the user is talking about cleaning doesn't seem _that_ unreasonable.
Flozzin · 1h ago
The article digs a little deeper after. Saying the chat records are lost, and that when they asked ChatGPT, it didn't give that guidance about cleaning purposely only, and that it never asked why they wanted to know.
Really though, this could have just as easily happened in a google search. It's not ChatGPT's fault as much as this persons fault for using a non-medical professional for medical guidance.
zahlman · 1h ago
>and that it never asked why they wanted to know.
Does ChatGPT ever ask the user, like, anything?
fl7305 · 33m ago
> Does ChatGPT ever ask the user, like, anything?
Yes. At least when I just tried ChatGPT-5:
Can I replace sodium chloride with sodium bromide?
ChatGPT said:
Yes, in some cases — but it depends on the application.
Chemistry/lab use: Both are salts of sodium and dissolve similarly, but bromide is more reactive in some contexts and heavier. It can change reaction outcomes, especially in halide-sensitive reactions (e.g., silver halide precipitation).
Food use: No — sodium bromide is toxic and not approved as a food additive.
Industrial processes: Sometimes interchangeable (e.g., certain brines, drilling fluids) if bromide’s different density, solubility, and cost are acceptable.
What’s your intended use?
zahlman · 1h ago
Wow. I thought this was just going to be about hyponatremia or something. (And from other research I've done, I really do think that on balance the US experts are recommending inappropriately low levels of sodium intake that are only appropriate for people who already have hypertension, and that typical North American dietary levels of sodium are just fine, really.) But replacing table salt with sodium bromide? Oof.
> to judge how badly chatgpt messed up, if there's no extra context assuming that the user is talking about cleaning doesn't seem _that_ unreasonable.
This would be a bizarre assumption for the simple reason that table salt is not normally used in cleaning.
OJFord · 1h ago
Yeah I thought it was a bit misleading too, it's not exactly 'stopping salt' that caused it, any more than you could describe the ill-effects of swapping nasturtiums for lily of the valley in your salads as 'avoiding edible flowers'.
kragen · 1h ago
He was poisoning himself for three months before he was treated, and apparently made a full recovery:
> He was tapered off risperidone before discharge and remained stable off medication at a check-in 2 weeks after discharge.
If you eliminated sodium chloride from your diet without replacing it with another sodium source, you would die in much less than three months; I think you'd be lucky to make it two weeks. You can't replace sodium with potassium or lithium or ammonium to the same degree that you can replace chloride with bromide.
OJFord · 5m ago
Even if you managed to reduce your intake enough to cause hyponatremia, I'm not sure that fits the 'rare condition' bill, and probably would've been discharged in well under 2 weeks after some fluids and advice.
Would be interesting if he started to become symptomatic and so asked ChatGPT and that's where he got the idea that it needed to be replaced with something though. (But I suspect it was more along the lines of salty taste without the NaCl intake.)
topaz0 · 1h ago
I'd say that the thing that messed up was the AI hype machine for pretending it might ever be a good idea to take chatgpt output as advice.
josefritzishere · 39m ago
This should be illegal. People are going to die because AI is too stupid for this responsibility.
> Should I replace sodium chloride with sodium bromide?
>> No. Sodium chloride (NaCl) and sodium bromide (NaBr) have different chemical and physiological properties... If your context is culinary or nutritional, do not substitute. If it is industrial or lab-based, match the compound to the intended reaction chemistry. What’s your use case?
Seems pretty solid and clear. I don't doubt that the user managed to confuse himself, but that's kind of silly to hold against ChatGPT. If I ask "how do I safely use coffee," the LLM responds reasonably, and the user interprets the response as saying it's safe to use freshly made hot coffee to give themself an enema, is that really something to hold against the LLM? Do we really want a world where, in response to any query, the LLM creates a long list of every conceivable thing not to do to avoid any legal liability?
There's also the question of base rates: how often do patients dangerously misinterpret human doctors' advice? Because they certainly do sometimes. Is that a fatal flaw in human doctors?
“Area man who had poor judgement ten years ago now has both poor judgement and access to chatbots”
Part of me rationalizes it as 'not exactly a discovery', which on its own was not a big issue before we were as connected as we, apparently, are ( even if I would argue that the connection is very ephemeral in nature ). I am still personally working through it, but at which point is the individual actually responsible?
I am not saying this lightly. I am not super pro-corporate, but the other end of this rope is not exactly fun times either. Where is the balance?
[1]https://theweek.com/articles/464674/8-drivers-who-blindly-fo...
still am, unless i re-interpret "do" as "due".
Your comment gave me a nightmare of that returning, but in AI form somehow.
If they are conditioned on a large data set that includes lots of examples of the result of people thinking, what they produce will look sort of like the results of thinking, but then if they were conditioned on a large data set of people repeating the same seven knock knock jokes over and over and over in some complex pattern (e.g. every third time, in French), what they produced will look like that, and nothing like thinking.
Failing to recognize this is going to get someone killed, if it hasn't already.
How can you so confidently proclaim that? Hinton and Ilya Sutskever certainly seem to think that LLMs do think. I'm not saying that you should accept what they say blindly due to their authority in the field, but their opinions should give your confidence some pause at least.
>How can you so confidently proclaim that?
Do you know why they're called 'models' by chance?
They're statistical, weighted models. They use statistical weights to predict the next token.
They don't think. They don't reason. Math, weights, and turtles all the way down. Calling anything an LLM does "thinking" or "reasoning" is incorrect. Calling any of this "AI" is even worse.
You are implicitly assuming that no statistical model acting on next-token prediction can, conditional on context, replicate all of the outputs that a human would give. This is a provably false claim, mathematically speaking, as human output under these assumptions would satisfy the conditions of Kolmogorov existence.
However, the status quo is that "AI" doesn't exist, computers only ever do exactly what they are programmed to do, and "thinking/reasoning" wasn't on the table.
I am not the one that needs to disprove the status quo.
Sure, they were evolved using criteria based on next token prediction. But you were also evolved, only using critera for higher reproduction.
So are you really thinking, or just trying to reproduce?
People believed in non existent WMDs and tens of thousands got killed. After that what happened ? Chimps with 3 inch brains feel super confident to run orgs and make decisions that effect entire populations and are never held accountable. Ask Snowden what happened after he recognized that.
> LLMs don't think. At all. They do next token prediction.
This is very often repeated by non-experts as a way to dismiss the capabilities of LLMs as some kind of a mirage. It would be so convenient if it were true. You have to define what 'think' means; once you do, you will find it more difficult to make such a statement. If you consider 'think' to be developing an internal representation of the query, drawing connections to other related concepts, and then checking your own answer, then there is significant empirical evidence to support high-performing LLMs do the first two, and one can make a good argument that test-time inference does a half-adequate, albeit inefficient, version of the latter. Whether LLMs will achieve human-level efficiency with these three things is another question entirely.
> If they are conditioned on a large data set of people repeating the same seven knock knock jokes over and over and over in some complex pattern (e.g. every third time, in French), what they produced will look like that, and nothing like thinking.
Absolutely, but this has little to do with your claim. If you narrow the data distribution, the model cannot develop appropriate language embeddings to do much of anything. You could even prove this mathematically with high probability statements.
> Failing to recognize this is going to get someone killed, if it hasn't already.
The real problem as in the article is that the LLM failed to intuit context, or to ask a followup. While a doctor would never have made this mistake, the doctor would know the relevant context since the patient came to see them in the first place. If you had a generic knowledgeable human acting as a resource bank that was asked the same question AND requested to provide nothing irrelevant, I can see a similar response being made. To me, the bigger issue is that there are consequences to easy access to esoteric information for the general public, and this would be reflected more in how we perform reinforcement learning to assert LLM behavior.
And then again, would you say that you cannot build a (presumably extremely complex) machine that thinks?
Do you think our brains are not complex biological machines?
Where I agree is that LLMs are absolutely not the endgame. They are super-human litterary prodiges. That's it. Litterary specialists, like poets, writers, scenarists, transcriptors, and so on. We should not ask them anything else.
Humans don't think. At all. They do next token prediction.
If they are [raised in environments] that includes lots of examples of the result of people thinking, what they produce will look sort of like the results of people thinking, but then if they were [raised in an environment] of people repeating the same seven knock knock jokes over and over and over in some complex pattern (e.g. every third time, in French), what they produced will look like that, and nothing like thinking.
I believe this can be observed in examples of feral children and accidental social isolation in childhood. It also explains the slow start but nearly exponential growth of knowledge within the history of human civilization.
I’m not going to hash out childhood development here because I’m not paid to post but if anyone read the above and was even slightly convinced I implore you to go read up on even the basics of early childhood development.
That's kind of like taking driving lessons in order to fix an engine. 'Early childhood development' is an emergent property of what could be cumulatively called a data set (everything the child has been exposed to).
No comments yet
Sure the machine very confidently lies about dangerous things, but you shouldn't trust it. But you should employ it to replace humans.
Plus, if you don’t have some completely obvious dread disease, doctors will essentially gaslight you.
These researchers get up on a pedestal, snicker at creative self-help, and ignore the systemic dysfunction that led to it.
>The patient told doctors that after reading about the negative effects of sodium chloride, or table salt, he consulted ChatGPT about eliminating chloride from his diet and started taking sodium bromide over a three-month period. This was despite reading that “chloride can be swapped with bromide, though likely for other purposes, such as cleaning”. Sodium bromide was used as a sedative in the early 20th century.
In any case, I feel like I really need to see the actual conversation itself to judge how badly chatgpt messed up, if there's no extra context assuming that the user is talking about cleaning doesn't seem _that_ unreasonable.
Really though, this could have just as easily happened in a google search. It's not ChatGPT's fault as much as this persons fault for using a non-medical professional for medical guidance.
Does ChatGPT ever ask the user, like, anything?
Yes. At least when I just tried ChatGPT-5:
Can I replace sodium chloride with sodium bromide?
ChatGPT said: Yes, in some cases — but it depends on the application.
Chemistry/lab use: Both are salts of sodium and dissolve similarly, but bromide is more reactive in some contexts and heavier. It can change reaction outcomes, especially in halide-sensitive reactions (e.g., silver halide precipitation).
Food use: No — sodium bromide is toxic and not approved as a food additive.
Industrial processes: Sometimes interchangeable (e.g., certain brines, drilling fluids) if bromide’s different density, solubility, and cost are acceptable.
What’s your intended use?
> to judge how badly chatgpt messed up, if there's no extra context assuming that the user is talking about cleaning doesn't seem _that_ unreasonable.
This would be a bizarre assumption for the simple reason that table salt is not normally used in cleaning.
> He was tapered off risperidone before discharge and remained stable off medication at a check-in 2 weeks after discharge.
https://www.acpjournals.org/doi/epdf/10.7326/aimcc.2024.1260
If you eliminated sodium chloride from your diet without replacing it with another sodium source, you would die in much less than three months; I think you'd be lucky to make it two weeks. You can't replace sodium with potassium or lithium or ammonium to the same degree that you can replace chloride with bromide.
Would be interesting if he started to become symptomatic and so asked ChatGPT and that's where he got the idea that it needed to be replaced with something though. (But I suspect it was more along the lines of salty taste without the NaCl intake.)