This is impressive in the worst way! There is a stark difference between original YC management and the current one. I just cannot believe someone who has such a limited understanding of the industry & tech can be the CEO of YC. Following a leader like Paul Graham is practically impossible (and rest of the founding team was surely very talented) but this is just horrible. Not sure if it matters whether he's good or bad in the short term, but after 5 years of this kind of management they'll absolutely lose their advantage.
It's ironic how YC became the Google/Microsoft of its industry.
Was it really just 4 years ago that NFTs were "the next big thing"?
The tweets don't age well in hindsight but sometimes there are technologies which feel like they might break through but never do. Having been bullish on the concept of NFTs doesn't make a strong argument for or against having good intuition of breakthrough innovation.
Why are we pretending like his comments about any of these things are as a neutral observer instead of as an investor cheerleading his investments? Why should anyone take Gary Tan seriously as a futurist?
Well, that's a good question but the answer isn't going to be to anybody's liking. Because he's got money. People equate having money with wisdom rather than with intelligence and intelligence is dual use, you can use it for good and you can use it for bad just as easily. It may lead to wisdom but that's fairly rare. Most of the time it just leads to money.
So people will follow those with money (or that they perceive to have money) without much critical thought about where that is going to lead them, they're hoping for wisdom but may end up being misled. That's why all of these ultra wealthy folk turned on a dime when the political weather changed, they don't really have principles, they just want more zeros.
No no, NFTs were ridiculously stupid, that's why it was so controversial, that's why there was so much backlash.
Tan having been bullish on NFTs is a very good indicator that he isn't hyping $TRENDING_THING based on technological merit
Having been bullish on NFTs, which were never more than hype in the primary use case promulgated for them, is absolutely a strong argument against having good intuition of breakthrough innovation. It demonstrates an inability to differentiate between hype and utility.
NFT's for real estate ownership, container tracking etc. could still have some form of utility. But what people think of when they hear NFT's isn't that, it's shitty monkey jpg's.
NFT's were never the next big thing, except for a very specific subset of very gullible idiots.
> Having been bullish on NFTs, which were never more than hype in the primary use case promulgated for them, is absolutely a strong argument against having good intuition of breakthrough innovation.
I always thought that NFTs were completely ridiculous and essentially nothing but hype. But then again, I thought that amazon wasn't going to work either, when I was there building it, so I'm not sure that even in a given individual "good intuition for breakthrough innovation" is a unitary thing.
That's a good point. I've had the same: I lost a bet on HN for a hundred bucks that Facebook would never break a billion in revenues. I - mistakenly - thought people would not be so stupid as to hand over their private lives to the likes of Mark Zuckerberg. But they did.
NFTs are just as stupid, if not more so and this time at least it looks like sanity prevailed. But the problem is more complex than just boolean 'made' or 'fail', and I think that's where the investment angle comes in. Investors bet on 'the next wave' all the time. And NFTs looked to the clueless as much as 'the next wave' as mobile phones or the transistor did at some point in time. The big differentiator to me is whether or not a thing like that requires a belief system or not. If it does then I don't give it much chance. But then we have all of crypto as a counterexample and quite a few people got stupidly rich peddling that.
NFTs followed the exact same path as crypto, which many predicted would happen. Crypto became a speculative asset traded on exchanges because it was too volatile and transaction-cost heavy to ever be used as a medium of exchange. NFTs being crypto-based were soon descended upon by finance bros and scammers who saw the opportunity for a quick buck, eliminating any possibility to develop it for utilitarian things like house deeds and concert tickets or whatever.
Yes. Blockchains in general, then 'metaverse', then NFTs, now AI. The hype-ier bits of the tech industry always need _something_ to be over-excited about.
(There was actually a short gap between the final collapse of the NFTs and ChatGPT; it's a wonder VCs were able to get out of bed in the morning)
> Having been bullish on the concept of NFTs doesn't make a strong argument for or against having good intuition of breakthrough innovation.
It makes a good argument that he is inclined to be overly impressed by whatever old nonsense people are currently pushing on twitter.
Vibe coding is more like the Visual Basic of this generation. It makes it much easier for less technical people to create software or for hackers to be much more productive, but there's still going to be a huge need for professional software development. It's not like everybody is going to become a vibe coder and there won't be a need for SaaS or low code solutions. I think tech people overestimate the capability and willingness for the average Joe to vibe code or engage with technology beyond the minimum required.
> Vibe coding is more like the Visual Basic of this generation. It makes it much easier for less technical people to create software or for hackers to be much more productive, but there's still going to be a huge need for professional software development [emphasis mine].
Visual Basic has never been well-regarded as a platform for "professional software development," so the analogy doesn't fit in that aspect.
The point I was making was VB was never considered a tool "for hackers to be much more productive," only a tool "for less technical people to create software."
I knew some hackers who used it back in the day. They were the types who were good with computers and could write shell scripts, but they were not professional programmers. They knew how to do what they needed done. The people I know today who are into vibe coding kind of fit that same mold. They want something done and they do it themselves, but they aren't necessarily good at coding or enjoy doing it.
Ehh, I dunno, it was really, really popular back then. I would bet that a non-trivial number of apps were built using VB by actual software engineers. Couldn't find concrete numbers but this article claims at its peak, 2/3rds of all business apps on Windows PCs were in VB: https://retool.com/visual-basic
I recall seeing inventory management systems, airline booking apps for travel agents, custom CRMs, internal LoB apps, check-in kiosks, vending machines, etc. with the tell-tale VB UI, especially the typical VB error dialog after a crash!
I messed around with several other UI toolkits of that era -- AWT, Swing, Qt, Flex/ActionScript -- and none was as productive as VB for simple apps. It was just the right amount of simplicity and development velocity for the myriad simple use-cases that were perfectly happy with rigid layouts.
This is perhaps what most non-dev people don't get. Maintenance is a far more harder thing than building something. So you want to go slow when building things, not fast. Either way building things fast has been a solved problem for a while, people don't go fast not because we don't have tools, but there are other fairly valid reasons to go slow. This is true with so many other things outside of software. I guess its called 'haste'.
This is true for most things. Especially where money and life are at stake. But Im guessing you could extend this to anything where reputation is at stake.
Im guessing it doesn't apply to some start ups, but other wise every one is subject to this.
Moderators didn't touch (or see) this until later, and the only way we intervened was to turn off the flags and flamewar penalties. HN users flagged it and the flamewar (or signal vs noise) detector pulled it well off the front page before moderators knew anything about it.
> I thought HN policy is to moderate less, not more, when YC is involved.
As dang wrote years ago [1], "Moderating HN less, however, does not mean not moderating at all—that would be a loophole you could drive a truck through".
But the real issue with this post is... there's nothing interesting or scandalous here, is there? Just someone (sure, our president) expressing opinions – which are all consistent with his business and investment activities.
By posting a narrative (that most likely went through some form of Strategic Comms team) comparing Zoho against a vibe coded product while also showcasing some of YC's star vibe-coded products, Garry is able to both craft a narrative that helps support YC's portfolio as well as bring a couple of people to start thinking about Vibecoding. It also acts as an indirect attack on G-Suite without calling out a massive organization like Alphabet by name, which YC needs to coexist with because a large portion of YC portfolio companies will either be acquired by Alphabet or will take or have taken some amount of funding from Alphabet and Alphabet related personal, and Alphabet personal are LPs in YC.
It doesn't matter if the take is right or wrong - it's started a discussion, and maybe one or two Zoho customers have now heard of a couple YC products to consider (outside of the HN bubble, very few people know about vibe coding or AI/ML).
All businesses do this, and knowing Zoho, they will probably leverage this as well as a way to market data sovereigninty and "make in India" by raising the specter of the big bad American capitalist trying to undermine a bootstrapped Indian company.
Anyhow, Zoho has built it's own foundation model [0] and offers an Agents marketplace for domain-specific agents [1]. I'm not sure if can compete head on against a LLaMa or DeepSeek on from sheer performance perspective, but it's good enough (something which a lot of engineers forget is more important than being perfect) to build a "data sovereignty" and "tech nationalism" story which they will absolutely run with as a result.
The only thing about this constant discussion about the metas and what not is that clearly everyone knows this, so who is this supposed to be for? I'm a peanut gallery member here, not a VC or a person who needs to appeal to VCs honestly, but then you can take my opinion as an unbiased third party... I could have guessed 1 and 2 are were motives for this kind of talk. But...clearly everyone else who actually cares about this game because it affects their livelihood or money is already aware VCs do 1 and 2, and unless you're exceptionally impressionable, it wouldn't work on you, no?
I have a theory but the primary one is not very flattering to those involved.
>The only thing about this constant discussion about the metas and what not is that clearly everyone knows this, so who is this supposed to be for?
LPs. Your LPs want to hear about how you are using their capital for A.I. If A.I fails 2 years down the line you will get to hide out in the crowd. Contrarianism in an institutional setting is actually very hard to do.
1. A number of media and substack articles about a Twitter beef that becomes a submarine article for Replit, Emergent Labs, and Taskade
2. A couple executives who may have not heard about these startups (this is actually very common outside the tech IC bubble) and will now ask their tech teams to contact them to see if they fit their needs
3. LPs who invested in YC by further reshoring Garry and YC's entire investment thesis.
> it wouldn't work on you...
It doesn't matter that it doesn't work on me - everyone does this form of narrative building (that's the entire point of Strategic Comms teams) so you have to play the game because that's how a Nash Equilibrium be.
Building a zeitgeist is a core part of demand gen, and yea it is transparent and tacky, but even 1 conversion for what was basically 1 hour of drafting makes the RoI positive.
It's for the press and the general public who don't really understand the tech. The hype can usually be sustained for around two years.
The press nearly always obliges in the first year. It works for wars and pandemics, too. In the second year the first dissenters begin to show up. We are now in past the stage of opinion reversal, where the press and the public mostly hate "AI".
This is also why it is wrong to compare "AI" to the early Internet. After the Internet bubble burst, the public still liked the Internet.
> This is also why it is wrong to compare "AI" to the early Internet. After the Internet bubble burst, the public still liked the Internet
The comparison is from a business perspective, and I absolutely stand by the comparison with the early generation of Internet companies with the early generation of AI companies.
Anyhow, the general public doesn't matter. It's capital (private and public) along with business cycles that create markets, and most of the public doesn't have the knowledge or the capital to make a difference.
It's capital that can force markets to exist but without the public to consume it's just wasted capital if value doesn't materialise.
The crux of the whole AI boom is exactly how much value it can materialise given the capital expenditure being absurdly astronomical, if it doesn't become the next trillion+ US$ market it will be a huge misallocation of capital.
The AI/ML products that are FCF positive and seeing significant traction are those that are complementing workflows in some shape or form by reducing
AI is now a loose used term that is encompassing 3 loosely connected markets that have now fallen under the same umbrella:
1. Construction/Land Speculation: a large portion of the AI story you hear about is a DC construction story
2. Hardware: a large portion of the AI story is just a rebranding around GPU fab and design, especially due to issues around subsidy disbursement under the CHIPS act
3. SaaS/Applications: a lot of products being derisively called as "LLM wrappers" are not cool from a technical perspective, but from an RoI perspective are good enough - $30k for data entry automation that is 80% right is cheaper than hiring a data entry team of 4 who cost $60k each.
Much of the bubble is due to 1, but 2 and 3 are somewhat insulated because of FCF and adjacent markets and narratives to pivot to (eg. For 2 it was "Chip Wars" 2-3 years ago and before that it was "ML" for 1-2 years and before that it was "Precision Medicine"...)
> without the public to consume...
The AI story really isn't a B2C story no matter how much people try to shoehorn it. The Cs are perpetually broke and margins are shit. The value that arises from automation like AI is around workflow and workforce augmentation in some shape or form, which makes it a B2B play.
https://x.com/garrytan/status/1506769562468958210
"If you own an NFT..." you are an innovator among innovators. - Garry Tan, 30 September 2021 [quote tweet]
https://x.com/garrytan/status/1443460589049704450
Web3 vs Earlier incarnations technology adoption curve [image tweet]
https://x.com/garrytan/status/1521530531568963584
"airdrops going mainstream could truly upend some % of the centralized ad-based economy of Web2" - Garry Tan, 25 Dec 2021
https://x.com/garrytan/status/1474826162408808448