Hacker News Reader

Comment by afavour | Hacker News Reader

I am well outside of this world so I'm probably just misinformed but when I think back to circa 2010 VC-backed startups felt like they were doing a wide array of interesting, diverse things. Today everything I see just reads like groupthink: crypto is a thing? OK, now _everything_ is crypto. AI is a thing? Hot damn, time to make everything AI-based!

Not sure when it happened but the tech world seems to be chasing its own tail. Maybe it's simply because the tech isn't so interesting these days. The release of the iPhone and Android was a cambrian explosion in possibility: the likes of Uber simply can't happen without a GPS enabled smartphone. But that's all old news now, so VCs are leaping on absolutely every tech advancement as the next big thing and then stumbling when it isn't.

View on HN

Your recollection is wrong. At any point in time, it has seemed like VCs and startups were chasing the same idea of the day. The joke in early 2010s was that every startup was "Uber for dogs."

Also, to your comment "tech isn't so interesting these days"... we get a new cambrian idea or core technology every 5-10 years, and obviously AI/ML is the current one since Nov 2022.

> we get a new cambrian idea or core technology every 5-10 years, and obviously AI/ML is the current one since Nov 2022.

Seems like a backwards way of looking at it. There is _always_ a cambrian explosion every 5-10 years therefore AI is that explosion? Seems like you're asking for false positives. What if we _don't_ have one every 5-10 years?

You're conflating an observation about the past with a prediction about the future. There's no reason to do that.

> You're conflating an observation about the past with a prediction about the future

No I'm not? OP said:

> we get a new cambrian idea or core technology every 5-10 years

I'm not the one conflating anything here.

> OP said

Exactly. Just as we also see the sun rise each morning.

That, however, does not imply one predicts that the sun will forever rise each morning. In fact, we know that at some point the sun will no longer rise in the morning. But we see the sun rise each morning, and we see a new cambrian idea or core technology every 5-10 years.

> I'm not the one conflating anything here.

That's fair. Conflation implies intent. Whereas you are just confused.

> You're just confused.

And you are being patronising.

My point is that there’s no guarantee this tech explosion will continue. Your logic appears to be “it has, so it will, until it doesn’t”, which… okay? All of your comments so far feel like you’re disagreeing with me just because you feel like being argumentative rather than because you have something useful to say.

It doesn't actually:

> Conflation is the merging of two or more sets of information, texts, ideas, or opinions into one, often in error.

It can be done as a rhetorical ploy, but that's not part of the definition, it can also simply be a mistake.

EXACTLY! It's like "AI" is here only because VC needed a next "thing."

ChatBots are super cool and useful but they are not the AI being sold.

AI is here because an AI product came out that millions of users found actually useful. The massive success of ChatGPT is what brought the current era of hype around AI - it's not anything VCs did.

Chatbots sucks, I want a human.

> What if we _don't_ have one every 5-10 years?

Then the observed trend will come to an end. Seems rather obvious, no?

I'm not afraid to say that AI/ML is not interesting. The last new tech that I found interesting was HTML5

I'm with you but to say text/image generation isn't interesting is just wrong. It is VERY interesting but selling it as "AI" is just wrong.

"interesting" is in the eye of the beholder. I simply don't find text generation that interesting, especially as a software engineer, and I REALLY don't find image generation interesting at all.

As a software engineer, you don’t have to.

But Hollywood does. Lots of producer/PA/VFX work is being rapidly automated and the genie is not going back inside the bottle.

that was never in question

Do you still find HTML5 to be as interesting as ever, or do you find that interest wanes with time? AI is many decades older than HTML5, so if your interest does wane then it stands to reason that there would be very little vestigial interest remaining.

I find WebGPU interesting as an extension of HTML5; I think HTML5 offers features that are not being taken advantage of by mainstream websites. I think HTML5 is interesting because, as a software engineer, it's a set of tools I can build something with. LLMs are not interesting to me because they're simply a product to use or consume. They're not terribly interesting to interface with because the commercial ones simply sit behind an HTTP API, so the only thing to do with them is send them different prompts, which is, again, uninteresting.

They did it in the 90s as well, and it worked out fabulously.

"The internet is a thing? Okay, now _everything_ is the internet."

They lean hard into new technologies. That's the nature of VC. Sometimes they're right and sometimes they're wrong.

It's not just VC. People are like that. There's a song with a sample "nothing gets old as fast as a new sound" that I have no idea where the sample was taken for further context. But it's pretty accurate and can be applied to anything. Clothing styles, home decor styles, art world fads, and on and on. People like to latch on, FOMO, whatever.

Seems a bit dramatic. I predict there is huge market for some data, that may require a bit of time for social acceptance. How about an app that alerts you when you are in a high crime area? Near a convicted violent offender? Like the person that is tailgating you even though there is no traffic around? Heat maps for income, taxes paid, crime, almost any type of data that can be tranched, bought and sold. 15 years ago the government had difficulty locating illegal residents after they entered the country. Now a lot of these people stand out like a blinking red light unless they only use cash and don't have any paperwork like a registered vehicle. There are a lot of companies that are seeking to monetize data in various ways. There has also been a resurgence for DOD nav/targeting data that works in a GPS-denied environment, due to systems reliant on GPS has proven to be a failure in Ukraine. The legendary $100k Excalibur shells that have intense demonstration videos are disabled by a simple GPS jammer.

Also keep in mind there is always a percentage of startups that exist only to be acquired by Microsoft/Oracle/Cisco and be shunted or thrown away/migrated to their products.

The Bay Area in general is anemic in general for financing. It has not and likely will not recover from the pandemic anytime soon. There has been hemorrhaging of companies and jobs. Oakland has a $90 million+ deficit this year and will likely need to cut the only thing they can, police and fire department employees. BART rider trips are only 43% from what they were in 2019. Leadership failure 101. Oakland mayor recalled. Alameda DA recalled. San Francisco mayor voted out. People hate bipping.

Early 2010 was people developing on mobile platforms. I remember 2010 was when Skype finally had video calls on mobile cellular.

I think 2014- was probably when things got more dynamic. Mobile was starting to mature, it was right before the crypto pump, iot was developing, and there were some big buy outs with whatsapp, oculus and viber.

Also, I’d say that was around the time the everything store of Amazon started gaining mainstream adoption.

The majority of 2010 startups were kids with no business experience banging out basic CRMs and marketplaces on Ruby. See for yourself: https://www.ycombinator.com/companies?batch=W10

Compare that to one of this year's YC batches, which IMO sounds much more sci-fi and interesting: https://www.ycombinator.com/companies?batch=W24

It seems like it's popular today to try to adopt a very salty and jaded attitude about everything with the idea that it's "calling BS" or bringing some kind of subversive viewpoint. In fact, nothing is more common, and it brings very little to the discussion.

>It seems like it's popular today to try to adopt a very salty and jaded attitude about everything with the idea that it's "calling BS" or bringing some kind of subversive viewpoint. In fact, nothing is more common, and it brings very little to the discussion.

You're right that it's common, but I don't think it's unwarranted. Over the last ~15 years, we've seen "amazing tech of the future" either work but have awful downstream effects (mainly social and economic), or flat-out not live up to the hype: social media, VR, NFTs, streaming media, the gig economic, etc.

It's healthy to tamp down hype and demand good thoughtful technological development, instead of just YOLOing into moving fast and breaking things.

Reply on news.ycombinator.com