[kwlug-disc] is AI profitable?

[Paul Nijjar]

At the risk of being a jerk (and no shade intended at Raymond or
anyone else), I am going to critique this AI response.  It is not
wrong but I feel it is misleading, especially in the "things to watch
out for".

No, this is not a "rigorous financial analysis platform." Was that the
claim? The site itself says: 

"All figures are estimated cumulative totals (all-time). Because most
of these companies are private, the numbers aren't exact; instead,
they are built from leaked financials, SEC filings, earnings calls,
and industry estimates from sources like Bloomberg, the WSJ, The
Information, and Epoch AI (all referenced at the bottom). The
punchline 'EVERYONE IS BROKE' is intentionally punchy, but shouldn't
be taken absolutely literally. "

That addresses the "transparent methodology, audited data sources, or
detailed assumptions" snipe. The methodology is not transparent but
comprehensible: the anonymous author links to sources for each of the
companies, and derives estimates for each of them based upon those
sources. We would have to double check that the bars on the graph
correspond to the information in those links, but at least we have the
ability to do that. (And who knows? Maybe this site is vibe coded and
full of hallucinations too.) 

It is not "likely" that these are estimates stitched together from
public reports. The website states this outright.

The equivocations about AI profitibility being "hard to measure" are
mostly bunk. The data sources are breathlessly reported profitability
numbers, often from the AI companies themselves. The point that the
companies are in growth mode is relevant to the question of whether
these companies OUGHT to be profitable at this point, but it has
nothing to do with whether these companies actually ARE profitable,
which is the question this website attempts to answer.

For capitalist companies, how is profitability "hard to measure"? It
may be difficult to obtain that information publicly, but are we to
believe that this information doesn't exist in the spreadsheets of the
AI companies, and that they are not carefully tracking this metric?

It may seem like nitpicking, but I find the worst insult in this
response the "directionally reasonable" bit. It is more than
directionally reasonable, but using that language conveys the idea
that this is a fun toy not based in truth, and that should not be
taken seriously. 

I agree fully with the assessment that this is a "simplified
visualization of AI spending vs returns". But I would strengthen this
to say that it is likely an accurate visualization, based upon
what public information is available. That matters. Should you be
cautious about this anonymous site that is on the Internet? Sure!
Should you be dismissive of it? Maybe not? 

I am sure I will regret making this critique (and I reiterate that I
am not attacking Raymond, or demanding a prohibition on AI responses
on the list), but I do think there is something misleading in the tone
of this response. The problem is that the response sounds so
reasonable, so a quick read leads one to dismiss the site.  But in
order to analyse and possibly refute its opinion, you have to spend
time digging deep and checking its work. 

I will also note that the response itself makes a bunch of claims
without citing sources. (eg "Even mainstream coverage has echoed that
framing.") I would not demand sources of a human being when
responding on a mailing list, but I feel the standard for these
chatbots should be different, because their output is not trustworthy.

- Paul

On Tue, May 26, 2026 at 09:33:22PM -0400, Raymond Chen wrote:
> I sent the link to AI and it responded:
> 
> I’d be cautious with  isaiprofitable.com.
> 
> From what I can tell, it’s more of a provocative “AI economics tracker”
> than a rigorous financial analysis platform. The core message — that most
> AI companies are spending huge amounts while only a few infrastructure
> players like NVIDIA are making large profits — is directionally reasonable.
> Even mainstream coverage has echoed that framing.
> 
> That said, there are a few things I’d watch for:
> 
>    - The site doesn’t appear to provide transparent methodology, audited
>    data sources, or detailed assumptions.
>    - The numbers are likely estimates stitched together from public
>    reports, capex disclosures, and speculation.
>    - “AI profitability” is extremely hard to measure because:
>       - AI revenue is often bundled into cloud/business units
>       - companies capitalize infrastructure differently
>       - R&D accounting varies
>       - many firms are intentionally in growth mode, not profit mode
> 
> So the site may be useful as:
> 
>    - a conversation starter
>    - a macro sentiment indicator
>    - a simplified visualization of AI spending vs returns
> 
> But I would not use it for:
> 
>    - investment decisions
>    - precise financial conclusions
>    - technical market analysis
> 
> One thing the site *does* get right conceptually: there’s a real
> distinction between:
> 
>    1. Infrastructure winners
>    Example: GPU vendors, cloud providers
>    2. Application-layer AI companies
>    Many are still subsidizing users heavily
> 
> A good real-world analogy:
> 
>    - During the gold rush, shovel sellers often made steadier money than
>    prospectors.
>    - In AI, NVIDIA selling GPUs has often been more profitable than AI apps
>    trying to monetize end users.
> 
> I also couldn’t find much independent reputation history or strong
> third-party validation around the site itself. The only broader mentions I
> found were articles discussing its claims, not validating its accuracy.
> 😁
> 
> On Tue, May 26, 2026 at 20:52 CrankyOldBugger <crankyoldbugger at gmail.com>
> wrote:
> 
> > Guy at work sent me this link:
> >
> > https://isaiprofitable.com/
> >
> >
> > _______________________________________________
> > kwlug-disc mailing list
> > To unsubscribe, send an email to kwlug-disc-leave at kwlug.org
> > with the subject "unsubscribe", or email
> > kwlug-disc-owner at kwlug.org to contact a human being.
> >

> _______________________________________________
> kwlug-disc mailing list
> To unsubscribe, send an email to kwlug-disc-leave at kwlug.org
> with the subject "unsubscribe", or email
> kwlug-disc-owner at kwlug.org to contact a human being.