| AI fears: ‘echoes of dot-com crash’ on 11:32 - Feb 6 with 1109 views | bsw72 |
That's Deutsche Bank - UBS research is different as are other groups. The current boom is not like the DotCom, therefore the outcome will be different and I doubt quite as dramatic.
The dot‑com bubble was classic market mania: huge retail and VC inflows into thousands of Internet startups, many with little or no revenue or viable business models but heavy on people and infrastructrue costs, driven by FOMO and media hype. When the funding dried up the public markets re‑rated those firms sharply, telecom and infrastructure over‑investment proved excessive, and many companies failed; but note the episode also left durable infrastructure and a few dominant survivors (Amazon, Google) that built long‑term value.
Today’s AI boom (not bubble at the moment, as you never know it's a bubble before it bursts) looks different. Much AI activity is embedded in profitable companies or delivers measurable efficiency/revenue gains and the market is concentrated around large cloud, chip and software incumbents. That doesn’t mean there isn’t froth on the periphery: many startups and products are over‑hyped, valuations for some early ventures look disconnected from fundamentals, and heavy capital commitments (chips, data centres, power) could become stranded if expectations fall.
However rather than a single, inevitable wipe‑out like 2000, the likeliest near‑term outcome is sectoral corrections and consolidation: pockets of speculative collapse alongside continued, productive long‑term value creation in core AI capabilities.
It's interesting that the startups back in the DotCom era were multi-staff and heavy on VC investment and built around the long term stratgegy/returns however the AI boom is allowing some VCs to back single person startups with as little as $20K in a very much fail fast approach and move on with minimal risk. |  | |  |
| AI fears: ‘echoes of dot-com crash’ on 11:34 - Feb 6 with 1099 views | _CliveBaker_ |
| AI fears: ‘echoes of dot-com crash’ on 11:32 - Feb 6 by bsw72 |
That's Deutsche Bank - UBS research is different as are other groups. The current boom is not like the DotCom, therefore the outcome will be different and I doubt quite as dramatic.
The dot‑com bubble was classic market mania: huge retail and VC inflows into thousands of Internet startups, many with little or no revenue or viable business models but heavy on people and infrastructrue costs, driven by FOMO and media hype. When the funding dried up the public markets re‑rated those firms sharply, telecom and infrastructure over‑investment proved excessive, and many companies failed; but note the episode also left durable infrastructure and a few dominant survivors (Amazon, Google) that built long‑term value.
Today’s AI boom (not bubble at the moment, as you never know it's a bubble before it bursts) looks different. Much AI activity is embedded in profitable companies or delivers measurable efficiency/revenue gains and the market is concentrated around large cloud, chip and software incumbents. That doesn’t mean there isn’t froth on the periphery: many startups and products are over‑hyped, valuations for some early ventures look disconnected from fundamentals, and heavy capital commitments (chips, data centres, power) could become stranded if expectations fall.
However rather than a single, inevitable wipe‑out like 2000, the likeliest near‑term outcome is sectoral corrections and consolidation: pockets of speculative collapse alongside continued, productive long‑term value creation in core AI capabilities.
It's interesting that the startups back in the DotCom era were multi-staff and heavy on VC investment and built around the long term stratgegy/returns however the AI boom is allowing some VCs to back single person startups with as little as $20K in a very much fail fast approach and move on with minimal risk. |
"valuations for some early ventures look disconnected from fundamentals". Aint that the truth, absolutely ludicrous in some instances.
Saw a business recently with about $10m revenue and negative EBITDA with a market cap of $1bn. Based off nothing but a dream at this point. |  | | |
| AI fears: ‘echoes of dot-com crash’ on 12:18 - Feb 6 with 999 views | vilanovablue |
| AI fears: ‘echoes of dot-com crash’ on 11:34 - Feb 6 by _CliveBaker_ |
"valuations for some early ventures look disconnected from fundamentals". Aint that the truth, absolutely ludicrous in some instances.
Saw a business recently with about $10m revenue and negative EBITDA with a market cap of $1bn. Based off nothing but a dream at this point. |
I think this is the main issue, insane levels of investment whilst at present the actual amount return is marginal at best at present. [Post edited 6 Feb 12:54]
|  | | |
| AI fears: ‘echoes of dot-com crash’ on 12:33 - Feb 6 with 966 views | StokieBlue |
| AI fears: ‘echoes of dot-com crash’ on 11:32 - Feb 6 by bsw72 |
That's Deutsche Bank - UBS research is different as are other groups. The current boom is not like the DotCom, therefore the outcome will be different and I doubt quite as dramatic.
The dot‑com bubble was classic market mania: huge retail and VC inflows into thousands of Internet startups, many with little or no revenue or viable business models but heavy on people and infrastructrue costs, driven by FOMO and media hype. When the funding dried up the public markets re‑rated those firms sharply, telecom and infrastructure over‑investment proved excessive, and many companies failed; but note the episode also left durable infrastructure and a few dominant survivors (Amazon, Google) that built long‑term value.
Today’s AI boom (not bubble at the moment, as you never know it's a bubble before it bursts) looks different. Much AI activity is embedded in profitable companies or delivers measurable efficiency/revenue gains and the market is concentrated around large cloud, chip and software incumbents. That doesn’t mean there isn’t froth on the periphery: many startups and products are over‑hyped, valuations for some early ventures look disconnected from fundamentals, and heavy capital commitments (chips, data centres, power) could become stranded if expectations fall.
However rather than a single, inevitable wipe‑out like 2000, the likeliest near‑term outcome is sectoral corrections and consolidation: pockets of speculative collapse alongside continued, productive long‑term value creation in core AI capabilities.
It's interesting that the startups back in the DotCom era were multi-staff and heavy on VC investment and built around the long term stratgegy/returns however the AI boom is allowing some VCs to back single person startups with as little as $20K in a very much fail fast approach and move on with minimal risk. |
There is a lot of concern about the circular nature of the AI investment. A lot of the same money is being passed between the same companies and whilst it may be fine it may also blow up if one element of that funding fails.
https://www.theguardian.com/te
SB |  | | |
| AI fears: ‘echoes of dot-com crash’ on 12:53 - Feb 6 with 930 views | DanTheMan |
| AI fears: ‘echoes of dot-com crash’ on 12:33 - Feb 6 by StokieBlue |
There is a lot of concern about the circular nature of the AI investment. A lot of the same money is being passed between the same companies and whilst it may be fine it may also blow up if one element of that funding fails.
https://www.theguardian.com/te
SB |
And of course, none of the AI companies (outside chip manufacturers) actually making any money off it. I'm not surprised the VC firms are starting to wonder when they'll start making any sort of revenue to justify the enormous spending. |  |
| |
| AI fears: ‘echoes of dot-com crash’ on 13:02 - Feb 6 with 894 views | StokieBlue |
| AI fears: ‘echoes of dot-com crash’ on 12:53 - Feb 6 by DanTheMan |
And of course, none of the AI companies (outside chip manufacturers) actually making any money off it. I'm not surprised the VC firms are starting to wonder when they'll start making any sort of revenue to justify the enormous spending. |
OpenAI are trying to shoehorn in adverts to the replies now but that seems a rather desperate way to get some numbers before an IPO so the owners can take some money.
SB | | | |
| AI fears: ‘echoes of dot-com crash’ on 13:15 - Feb 6 with 860 views | Daninthecampo |
Its more a wobble than a bubble if you compare Cisco's growth during the dot.com era to that of Nvidia, totally different scenario, Nvidia surging earnings growth means the p/e is now lower whereas Cisco p/e vastly outpaced earnings growth
.
Its also expected Big tech is going to spend considerably more than anticipated on AI as it has for the last 2 years, 2025 anticipated was 22% growth when in actually was 68%.
As with all things lots of research before investing your money! |  | | |
| AI fears: ‘echoes of dot-com crash’ on 13:15 - Feb 6 with 857 views | Swansea_Blue |
| AI fears: ‘echoes of dot-com crash’ on 11:32 - Feb 6 by bsw72 |
That's Deutsche Bank - UBS research is different as are other groups. The current boom is not like the DotCom, therefore the outcome will be different and I doubt quite as dramatic.
The dot‑com bubble was classic market mania: huge retail and VC inflows into thousands of Internet startups, many with little or no revenue or viable business models but heavy on people and infrastructrue costs, driven by FOMO and media hype. When the funding dried up the public markets re‑rated those firms sharply, telecom and infrastructure over‑investment proved excessive, and many companies failed; but note the episode also left durable infrastructure and a few dominant survivors (Amazon, Google) that built long‑term value.
Today’s AI boom (not bubble at the moment, as you never know it's a bubble before it bursts) looks different. Much AI activity is embedded in profitable companies or delivers measurable efficiency/revenue gains and the market is concentrated around large cloud, chip and software incumbents. That doesn’t mean there isn’t froth on the periphery: many startups and products are over‑hyped, valuations for some early ventures look disconnected from fundamentals, and heavy capital commitments (chips, data centres, power) could become stranded if expectations fall.
However rather than a single, inevitable wipe‑out like 2000, the likeliest near‑term outcome is sectoral corrections and consolidation: pockets of speculative collapse alongside continued, productive long‑term value creation in core AI capabilities.
It's interesting that the startups back in the DotCom era were multi-staff and heavy on VC investment and built around the long term stratgegy/returns however the AI boom is allowing some VCs to back single person startups with as little as $20K in a very much fail fast approach and move on with minimal risk. |
Thank you. Interesting background around the key differences. Re. Stokies point, I’d noticed moves to monetarise AI output with adverts. I’ve no doubt AI will provide fantastic opportunities in some specialist fields, but for the rest of us we’re going to be drowning in cheap AI slop. The two combining is going to be a problem too unless the Advertising Standatds Authority wakes up. There’s already a lot of misleading AI adverts on YouTube. The scope for disinformation around events and political campaigns is huge too. [Post edited 6 Feb 13:45]
|  |
| |
Login to get fewer ads
| AI fears: ‘echoes of dot-com crash’ on 13:51 - Feb 6 with 766 views | bsw72 |
| AI fears: ‘echoes of dot-com crash’ on 12:53 - Feb 6 by DanTheMan |
And of course, none of the AI companies (outside chip manufacturers) actually making any money off it. I'm not surprised the VC firms are starting to wonder when they'll start making any sort of revenue to justify the enormous spending. |
That’s too sweeping. Many companies outside chipmakers are already monetising AI often in ways that don’t look like a separate “AI line” on the profit & loss.
Microsoft and Google earn huge sums from AI‑enhanced search, cloud and productivity services (Copilot, Gemini, Azure/GCP AI offerings); OpenAI, Anthropic and others sell paid APIs and enterprise licences; Adobe and Salesforce bundle AI features into paid Creative Cloud and CRM subscriptions; cloud providers (AWS, Azure) charge for managed inference and model hosting. Most of the cloud companies supporting AI infrastructure are already cash rich, hence they can buy the hardware (driving nVidia up) to further reinvest with less reliance on further debt funding.
Equally important, many firms are using AI effecgtively to lift conversion, cut cost-to-serve, reduce churn or speed R&D measurable economic benefits that translate into revenue or margin improvements even if they’re not labelled “AI revenue.” | | | |
| AI fears: ‘echoes of dot-com crash’ on 14:05 - Feb 6 with 705 views | WicklowBlue |
|  | | |
| AI fears: ‘echoes of dot-com crash’ on 14:24 - Feb 6 with 657 views | DanTheMan |
| AI fears: ‘echoes of dot-com crash’ on 13:51 - Feb 6 by bsw72 |
That’s too sweeping. Many companies outside chipmakers are already monetising AI often in ways that don’t look like a separate “AI line” on the profit & loss.
Microsoft and Google earn huge sums from AI‑enhanced search, cloud and productivity services (Copilot, Gemini, Azure/GCP AI offerings); OpenAI, Anthropic and others sell paid APIs and enterprise licences; Adobe and Salesforce bundle AI features into paid Creative Cloud and CRM subscriptions; cloud providers (AWS, Azure) charge for managed inference and model hosting. Most of the cloud companies supporting AI infrastructure are already cash rich, hence they can buy the hardware (driving nVidia up) to further reinvest with less reliance on further debt funding.
Equally important, many firms are using AI effecgtively to lift conversion, cut cost-to-serve, reduce churn or speed R&D measurable economic benefits that translate into revenue or margin improvements even if they’re not labelled “AI revenue.” |
Yes they are using these things but they aren't actually making any money off it.
There are only three companies actually make their own models with any real usage - Google, OpenAI and Anthropic. Not one of them is making money off them.
Sure, it's being bundled into everything and forced down people's throats but right now they cost waaaay more to run than they actually do to create revenue.
Right now I just fail to see how the sums add up. | |
| |
| AI fears: ‘echoes of dot-com crash’ on 14:56 - Feb 6 with 610 views | WicklowBlue |
| AI fears: ‘echoes of dot-com crash’ on 14:24 - Feb 6 by DanTheMan |
Yes they are using these things but they aren't actually making any money off it.
There are only three companies actually make their own models with any real usage - Google, OpenAI and Anthropic. Not one of them is making money off them.
Sure, it's being bundled into everything and forced down people's throats but right now they cost waaaay more to run than they actually do to create revenue.
Right now I just fail to see how the sums add up. |
Hard to see this as anything than a bubble to be frank. Billions being poured into AI, chips, data centres etc with no profit today being realised.
At some point a squeeze on that investment/liquidity will happen and mass contagion will break out with smaller startups being gobbled up by the big guys. In turn crystalising investment losses that will further spook markets and accelerate the cycle.
The only question for me is when and how far does this go. | | | |
| AI fears: ‘echoes of dot-com crash’ on 15:42 - Feb 6 with 548 views | bsw72 |
| AI fears: ‘echoes of dot-com crash’ on 14:24 - Feb 6 by DanTheMan |
Yes they are using these things but they aren't actually making any money off it.
There are only three companies actually make their own models with any real usage - Google, OpenAI and Anthropic. Not one of them is making money off them.
Sure, it's being bundled into everything and forced down people's throats but right now they cost waaaay more to run than they actually do to create revenue.
Right now I just fail to see how the sums add up. |
That's not true.
While the AI infrastructure pushes up costs in the short term, most of that spending is capital investment and written down over several years, so the yearly depreciation charge is smaller than the total build‑out.
That means AI‑driven revenue growth is already outpacing the annualised capital charge, and once utilization and pricing settle, the investment starts to deliver real, positive returns.
For example, Google revenue from AI services, including Google Cloud, grew by 29% in Q2 2025 alone, contributing to a total revenue of $96.43 billion. When you consider that Google has projected a capital expenditure of $85 billion for 2025, focusing on AI infrastructure and cloud computing capabilities - they are obviously making money off it. | | | |
| AI fears: ‘echoes of dot-com crash’ on 15:45 - Feb 6 with 519 views | DanTheMan |
| AI fears: ‘echoes of dot-com crash’ on 15:42 - Feb 6 by bsw72 |
That's not true.
While the AI infrastructure pushes up costs in the short term, most of that spending is capital investment and written down over several years, so the yearly depreciation charge is smaller than the total build‑out.
That means AI‑driven revenue growth is already outpacing the annualised capital charge, and once utilization and pricing settle, the investment starts to deliver real, positive returns.
For example, Google revenue from AI services, including Google Cloud, grew by 29% in Q2 2025 alone, contributing to a total revenue of $96.43 billion. When you consider that Google has projected a capital expenditure of $85 billion for 2025, focusing on AI infrastructure and cloud computing capabilities - they are obviously making money off it. |
You do realise that graphics cards that drive all of this spending only last a few years right? They need to be constantly replaced.
And your Google revenues are for all Google services so there is no way to know if AI is costing them money or not. You say it's "obvious" but I completely disagree. OpenAI are losing money hand over fist. So are Anthropic. [Post edited 6 Feb 15:46]
| |
| |
| AI fears: ‘echoes of dot-com crash’ on 15:48 - Feb 6 with 502 views | coote |
| AI fears: ‘echoes of dot-com crash’ on 15:42 - Feb 6 by bsw72 |
That's not true.
While the AI infrastructure pushes up costs in the short term, most of that spending is capital investment and written down over several years, so the yearly depreciation charge is smaller than the total build‑out.
That means AI‑driven revenue growth is already outpacing the annualised capital charge, and once utilization and pricing settle, the investment starts to deliver real, positive returns.
For example, Google revenue from AI services, including Google Cloud, grew by 29% in Q2 2025 alone, contributing to a total revenue of $96.43 billion. When you consider that Google has projected a capital expenditure of $85 billion for 2025, focusing on AI infrastructure and cloud computing capabilities - they are obviously making money off it. |
What are your thoughts on ‘saas’ companies?
My portfolio has taken an absolute hammering this year.
Better day today though.(amazon aside) |  | | |
| AI fears: ‘echoes of dot-com crash’ on 19:26 - Feb 6 with 394 views | stonojnr |
| AI fears: ‘echoes of dot-com crash’ on 15:48 - Feb 6 by coote |
What are your thoughts on ‘saas’ companies?
My portfolio has taken an absolute hammering this year.
Better day today though.(amazon aside) |
Saas, well lucky the club isnt sponsored by one isnt it 😉 | | | |
| AI fears: ‘echoes of dot-com crash’ on 19:40 - Feb 6 with 361 views | Camul123 |
Half the responses on this thread read as though drafted by AI ... | | | |
| AI fears: ‘echoes of dot-com crash’ on 19:46 - Feb 6 with 349 views | thebooks |
| AI fears: ‘echoes of dot-com crash’ on 11:32 - Feb 6 by bsw72 |
That's Deutsche Bank - UBS research is different as are other groups. The current boom is not like the DotCom, therefore the outcome will be different and I doubt quite as dramatic.
The dot‑com bubble was classic market mania: huge retail and VC inflows into thousands of Internet startups, many with little or no revenue or viable business models but heavy on people and infrastructrue costs, driven by FOMO and media hype. When the funding dried up the public markets re‑rated those firms sharply, telecom and infrastructure over‑investment proved excessive, and many companies failed; but note the episode also left durable infrastructure and a few dominant survivors (Amazon, Google) that built long‑term value.
Today’s AI boom (not bubble at the moment, as you never know it's a bubble before it bursts) looks different. Much AI activity is embedded in profitable companies or delivers measurable efficiency/revenue gains and the market is concentrated around large cloud, chip and software incumbents. That doesn’t mean there isn’t froth on the periphery: many startups and products are over‑hyped, valuations for some early ventures look disconnected from fundamentals, and heavy capital commitments (chips, data centres, power) could become stranded if expectations fall.
However rather than a single, inevitable wipe‑out like 2000, the likeliest near‑term outcome is sectoral corrections and consolidation: pockets of speculative collapse alongside continued, productive long‑term value creation in core AI capabilities.
It's interesting that the startups back in the DotCom era were multi-staff and heavy on VC investment and built around the long term stratgegy/returns however the AI boom is allowing some VCs to back single person startups with as little as $20K in a very much fail fast approach and move on with minimal risk. |
It’s a circle jerk of software companies hyping something that will never happen, agreeing investment deals with hardware providers, and then pumping the investment back into hardware providers. Smaller companies are riding on the wave of this massive grift. It will fail on a huge scale.
The only question is who will pick up the bill.
Incidentally, just another bit of weirdness in New New Labour. Why put so much faith in something that a) hasn’t actually delivered much and b) explicitly replaces workers and c) is funded and run by techno-fascists?
Larry Elliot has a good article on what the UK should be investing in, but far too old-fashioned no doubt:
https://www.theguardian.com/co |  | | |
| |
By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions.
