Investment forum flagging discussion and post-mortem

Development Ideas
2849

I never would have guessed that Juurikki, who is generally critical of AI, would end up defending the use of AI as a tool for information retrieval.

Who knows what the term “Non-Market Economy, NME” means? Hands up. Not many hands to be seen. In the entire history of the Inderes forum, the term has not been used once, according to the search engine, until Juurikki introduced it today, believing it was necessary for assessing country risk. Case: PYN Elite / Vietnam.

Having avoided business schools, Juurikki’s understanding of the term’s precise meaning was quite hazy until asking Google’s AI. It provided a neat summary, just like that.

Scientific argumentation is not required on the Inderes forum, but no one with a researcher’s training easily deviates from a few basic principles. The source is always cited, and if it is a direct quote, quotation marks are used. It would be dishonest not to disclose, when introducing a subject that is strange and largely new even to oneself, that it is an AI-generated summary.

At least Juurikki’s time is not sufficient for constructing very deep essays, when it is, after all, just a new deepening perspective on country risk assessment in a single discussion thread.

The Inderes forum serves as a school for Juurikki at least, where one can learn from those smarter than oneself and hopefully give something back in return. AI summaries can be useful for particularly specific questions, as long as they are used only when needed. Of course, one shouldn’t spam the forum with common knowledge, even if the AI’s summary is personally pleasing.

So, what does Non-Market Economy mean? Go check the PYN thread.

11 Likes
2850

I disagree with this. In my opinion, it is valuable to post relevant or interesting news and current affairs such as earnings reports, stock exchange releases, management interviews, analyst updates, financial media news, etc., to company threads with a relatively low threshold.

It is often even better to share such things neutrally first, and only then, if appropriate, add one’s own thoughts and opinions, so that opinions and one’s own conclusions do not get too mixed up with facts or other people’s opinions and conclusions. (As long as one doesn’t just dump some random X-link or AI-sloppiness or similar.)

Of course, the forum’s main added value comes from the participants’ own thoughts and well-reasoned writings, but this does not happen in a vacuum, and the forum’s function as a news aggregator is also important.

15 Likes
2851

Good point, and I don’t mind news being posted. But it would be nice if people could take the effort to include at least a short “Nokia filed for bankruptcy” intro instead of just a link. It takes 3 seconds. If someone needs a real-time news service, a discussion forum is not the right place for it.

8 Likes
2852

AI is an absolutely brilliant tool for information retrieval, company analysis, generating investment ideas, summarizing large amounts of data, explaining abbreviations and concepts, developing investment tools, etc. I don’t think anyone is questioning that. But when people discuss things here, I believe most want to talk with humans and not AI. You don’t even need this forum for the latter.

My original message was prompted by the fact that someone posted two rather long AI outputs as-is into a company thread. Without any original content. Without prompts. Pretty much zero content. And apparently, the mods agreed, as those two messages disappeared from there quickly.

A thread for sharing the best ways to utilize AI would be more useful. Then everyone could use those skills themselves and pull similar analyses for every company every single day.

16 Likes
2853

I posted the following message in the “wondering about share price changes” thread, which was flagged and removed:

It’s honestly beyond me what was so objectionable about this!

3 Likes
2854

Hi!

It wasn’t there; it was in the Stock Exchange thread and was flagged. I moved it to the Share Price Development thread after it was flagged. :slight_smile: But then, soon after my action, someone else must have flagged it, and the view probably hadn’t updated for the flagger or my colleague—meaning it likely still appeared to them as being in the Stock Exchange thread, so they deleted it. I restored it a moment ago after reading your message. :slight_smile:

Sometimes it happens that a message has already been deleted (by us or the author) or moved, but for some people, it hasn’t “updated” yet (it’s still visible to them), perhaps because they had that specific thread open or something (?), as some people have replied to messages that were deleted a while ago. :slight_smile:

Yeah, no worries. :slight_smile: And for the next flagger and moderator, it still appeared “incorrectly” even though it had been moved. :slight_smile:

I sometimes post in the wrong thread myself, for example, the wrong Valmet thread. :smiley:

6 Likes
2855

If that’s the case, then it ended up in the wrong thread by mistake. I specifically tried to choose a thread where it wouldn’t get flagged.

1 Like
2856

In the thread about earnings-related pension accrual, I asked for more details regarding the “hell” that the boomers supposedly went through. The message was flagged :smiley:

Apparently, for some, living through decades of continuous economic growth is hell; go figure.

1 Like
2857

I flagged this because the topic of the thread is Earnings-related pension accrual, as indicated by the title. There is no reason to go off on tangents, especially with one-line comments. You should write well-reasoned opinions according to the topic rather than nitpicking individual words in someone’s message. You should also learn to reply to comments in other threads; for instance, in the “Kahvihuone” (Coffee Room) section, the discussion is quite free-form.

11 Likes
2859

I recommend avoiding the thread in question anyway; its content seems very stale and one-sided whenever I’ve happened to stumble upon it.

3 Likes
2860

A tip for @amokk: there is an “Inderes’ failed recommendations” thread on the forum where you can post messages regarding recommendations from Inderes (the analysts) that went wrong. The “Coffee Room” thread is also a fairly general-purpose place for all kinds of messages.

Company-specific threads are intended for messages discussing those companies as investment targets.

There is a bit of a difference in thread culture compared to, for example, the Kauppalehti discussion forum. By following these guidelines, you will mostly avoid flags/message deletions and can make your voice heard here without experiences of censorship :+1:

11 Likes
Inderesin kahvihuone (Osa 11)
2862

In the Revenio thread, I posted an analysis of a press release generated using AI. The result:

“Your post was marked as spam: it is perceived as being promotional in nature and not a useful or appropriate addition to the discussion.”

The name of the AI was not mentioned, only that it had been used. If the AI had remained unmentioned, as is common practice, the text would have been left alone. Many texts here are identifiable as AI-generated. Of course, every investor nowadays uses all available tools, but the analysis must be done by everyone themselves.

2863

Let’s start by clarifying that I am not the person who flagged your message.

I saw your message, but skipped it immediately as it was AI-generated.

There has been so much discussion about this lately that AI-generated messages don’t really add any value to the forum (the flags in this case prove it). I won’t take a stance on whether many other messages are even worse, or how accurate the AI-generated message is.

11 Likes
2864

I strongly agree. Adding AI analyses is like posting a screenshot of your Google search results or a quote of an analysis provided by your neighbor Pentti whom you barely know. They do not bring any real added value to the discussion. Anyone could have a widely used AI generate one and paste it into the thread. There is no guarantee that the interpretations provided by the AI are correct or valid; they are just the AI’s current best guess at a correct “analysis.”

12 Likes
2865

Likewise. The point was the acknowledgment that AI has been involved. Of course, I read and accepted that analysis, which I thought was spot on. In my opinion, it should always be noted if AI has been involved, even partially.

3 Likes
2866

I think we should start a separate thread for all things AI, where people could freely post AI analyses and other AI-generated content. You could then just include a link to the longer post found in the AI thread in the relevant topic thread, so those interested in the subject could continue the discussion and interpretation there.

AI is certainly the future, and grumpy old folks like me will get left behind. Still, I would like to keep discussion forums dedicated to human-to-human discussion, for as long as that is possible.

1 Like
2867

Personally, it occurs to me that if all questions generated by AI are left unpublished, we might miss out on asking perfectly valid questions, meaning they won’t reach the general public even if they are worth considering.

Personally, however, I mostly just give AI-generated answers a cursory review.

2 Likes
2868

Could the Inderes forum moderators develop an AI agent 007 with the data department :wink: that reveals all posts where AI has been used? Thanks in advance.

Using AI wouldn’t be a sufficient criterion for rejecting a post, but it would be good to see a clear visual AI warning before starting to read the message. Universities already use plagiarism detection systems, so by adjusting those, you could probably get a system that works well enough—and cheaply, if you assign the coding to the AI itself.

And so that this development idea doesn’t trigger a flag, the post must be justified in its conclusion. Marking AI-assisted posts with a color code at the top of the message would reduce the need for flagging, as everyone could choose for themselves whether or not to read AI-assisted content.

5 Likes
2869

Removing hallucinating, thoughtless technology with the help of hallucinating, thoughtless technology. :thinking:

3 Likes
2872

Judging by the flags, sharing Jarkko Tontti’s column on this forum is against the rules. I am outraged!!!