His actual statement was much less dramatic than the headline gives it credit for.
That wouldn’t surprise me. I think you see a lot of hiring and layoff type decisions based on expectations of future needs rather than current needs.
1 Like
True, but AI is such a weird precedent
You’re projecting head count, efficiencies, and even expenses… off of technology that doesn’t theoretically exist.
It’s a combination of speculation (internally), and a combination of attempts to increase your share price because you’re publicly announcing AI tech (that doesn’t exist or isn’t implemented) will suddenly change your company
Why is it weird, you probably didn’t think the internet or the pc was going to be real either.
The internet and PC’s were tangible, material technology.
AI is not tangible, it’s speculative.
I would compared AI more to the dot com bubble.
1 Like
That doesn’t make much sense. For one, PCs are tangible but the internet isn’t. Also the internet and dot com bubble are very related.
Maybe you can say that AI is following the trajectory of the early internet towards a dot com bubble.
It’s very possible there is an AI bubble (like we saw with early internet), but also very possible (even likely) that AI will eventually have a similar or even larger impact than the internet despite that.
I meant that PCs & the internet (technology) were already conceptually developed when they were commercially scaled.
The Dot Com Bubble was birthed out of the internet, based on speculative internet stocks that were highly overvalued and showed no returns except for the lucky few that became successful.
True artificial intelligence (to where it’s profitable) hasn’t conceptually developed. This won’t happen until 2 things essentially happen:
- AGI technologically is developed (it doesn’t exist right now)
- The energy costs of data centers aren’t eating away at “AI” / Tech company balance sheets anymore [hence why so many Tech companies having drifted away from ESG, etc…]
Once AGI is developed, then there will be a material/tangible technology that actually has the capability of replacing humans at scale.
Right now, you have many companies [both tech and non-tech] who are projecting their financials on technology that doesn’t actually exist, but rather, there’s expectation that it will exist. It should also be noted that COVID produced over-hiring across numerous industries as well, so that’s also playing a factor in layoffs.
Anything else is really just automation that’s being labeled as “AI”. For example, automating service jobs (drive-thrus, cashiers, etc.) or analyzing documents in white collar work.
Currently, what’s touted as AI are really just more complicated chatbots (Large Language Models). That’s not to take anything away from it, but LLMs as a sole technology is not profitable unless it’s scaled and implemented commercially, which it has to do.
Only companies such as Google or Microsoft can implement LLMs because it relies on their search engines as hedge (Google Search/Bing Search)
Man, you are so misinformed. I lived through both and I’m seeing the same thing with AI. In the beginning there were lots of companies/players and the market chose just a few. Do you guys remember Compaq, Lycos, Netscape ? That’s what’s happening with AI, lots of players but in the end there will only be a few. The best ones, and we will be the ones that benefit from it.
1 Like
Compaq was bought by a company that essentially sold the same products/services. Same with Lycos and Netscape. They all had products that existed.
OpenAI, Gemeni, Meta AI, etc… are essentially all the same. They are Large Lanugage Models that uses the internet (open domain) to retrieve information. This technology, as a standalone, is unprofitable.
The difference between OpenAI and the other 2 companies mentioned is that they have balance sheets that are a hedge against the LLM product. OpenAI, does not have this.
As I mentioned above, companies (including non-tech) are already attempting to project efficiencies, financials, headcounts, etc… based on technology that doesn’t exist.
Once the AI Bubble bursts (if it does), then companies like Open AI will no longer exists or they will sale to a company like Microsoft. However, the problem with the ladder is that if nobody is investing into “AI” anymore, then companies like Microsoft will likely get rid of the technology itself because its too expensive to run without outside capital.
As it stands right now, Large Language Models are not generating profits to investors. What investors are waiting for is a breakthrough - / - AGI
Right now, AI is free for most users, it’s like when YouTube or google first started. Now you have to watch ads, but millions of people are using AI right now. So you are wrong. Its so rampant that schools have to try and devise way to thwart the use of AI. So keep saying AI isn’t real or isn’t there yet. I get it, you’re young and to you this type of technology always existed. But you weren’t there when they first started. You just think oh if the technology is good everyone will have it overnight. Wrong.
1 Like
First, we were talking about the hiring and firing decisions related to AI speculation. I believe (though I haven’t looked into this) many companies using AI to get efficiencies are using it from 3rd party providers. Thus it is a tool that they are leveraging for profit and affects their need of perceived needs for employee count. In this way, it is totally separate from any AI bubble with the companies actually producing AI infrastructure and the llm.
So when you’re talking about AI being unprofitable currently, you have to ask who is it unprofitable for? The developers or the users?
The developers are likely not going to be doing hiring or firing based on AI efficiencies, but would be doing so to get their own business into profitability. The user companies would be the ones who may hire or fire based on AI efficiencies.
Well would you look at that… eh?
Right on the cusp of this conversation, from JP Morgan themselves:
The Worst-Case Scenario for AI Would Mirror Dot-Com Bust, JPM Says - Business Insider
Investors.
LLM’s are still in the venture capital stage because “AI” Companies are selling their product based on non-exsistent potential → Reaching AGI.
The entire hype machine surrounding OpenAI was that the company was going to reach AGI, and the ChatGPT technology as we know it is merely a bridge to get there. Yet, there is no indication whatsover that OpenAI (or any tech company in general) is anywhere remotely close to reaching AGI.
AGI is what will turn AI into a commercial product. Until then, ChatGPT is merely a form of a Google Search. Still incredible technology, but in its current form, it’s not sustainable without continuous investment (for which investors are not receiving returns).
With all that I just said in mind, companies (tech and non-tech) are making corporate decisions based on tech that doesn’t exist because they’re trying to increase their share price by any means possible.
People have been saying that for months. That comparison has been widely discussed since early 2024.
Does anyone think it’s not the worst case scenario for stocks?
1 Like
I think it’s fair to say AGI is not real yet; but other aspects of AI are real today
and profitable.
I really not sure about all this. Advertising is the future it seems for the big ones.
And while LLM are not there yet…
No, most companies developing large language models (LLMs) are not currently profitable
, as they are in a massive “lose money to win market share” phase. High costs for training and running models are subsidized by venture capital and big tech investments, while they charge consumers less than the cost of service to encourage usage. Profitability is expected to be achieved in the future through controlling infrastructure, monetizing ecosystems, or owning niche markets, but many startups may not survive the “gold rush”.
Chatbots OTOH seem to be already there and delivering.
Yes,
chatbots can be highly profitable for businesses by significantly reducing operational costs and driving revenue growth. The global chatbot market is rapidly expanding, demonstrating the clear business value of these solutions.
Key Drivers of Chatbot Profitability
- Cost Savings: Chatbots can reduce customer service costs by 40-60% for enterprises. They automate repetitive tasks, handle a high volume of routine inquiries 24/7, and free up human agents to focus on more complex issues, leading to substantial savings in labor and operational expenses. Businesses have reported saving an average of $300,000 annually through chatbot use.
- Revenue Generation: Chatbots contribute directly to sales and marketing efforts.
- Lead Generation: 55% of companies using chatbots experience an increase in high-quality leads.
- Sales Conversion: Businesses have reported a 67% increase in sales through the assistance of chatbots, and some sales transactions see conversion rates as high as 70% in specific industries.
- Personalization and Upselling: They provide personalized product recommendations and assist with order tracking, which can increase annual revenue by 7-25% for e-commerce businesses.
- Enhanced Customer Experience (CX): Chatbots offer instant, 24/7 support, which meets customer expectations for immediate service and increases satisfaction and retention rates. Faster complaint resolution (90% of businesses witness this) and improved response times enhance brand loyalty.
- Scalability and Efficiency: Chatbot solutions are highly scalable, allowing businesses to handle increased customer demand without a linear increase in staffing costs.
II still think in terms of the hiring and firing decisions around AI usage in industrires is a different conversation than the AI bubble and if AI companies can be profitable. As I said before, many companies are leveraging AI but paying for the service rather than developing their own AI or LLM. So you really have two buckets of companies in that sense and generally the economics of the two buckets (one developing ai in house, one buying ai service like software) may have different outcomes and different decisions.
Not quite sure why you tagged me on that. I don’t disagree that there very well could be an AI bubble that will burst.
That means the surviving 60% will dominate the landscape. Just as with any free market segments. It’s proven. Just like the often mentioned dotcom bubble, it only made the strongest companies dominate, but didn’t eliminate the internet.
1 Like