lazy founder

grinding 9-5 to my first billion. or million. or until i run out of money.

[lazy founder @FounderLazy]

@GRITCULT yeah and then it takes me half an hour to peel the sweaty cotton shirt off my back

[lazy founder @FounderLazy]

@coldhealing "reserved" in this case just means you have to eat to get a table, they won't seat you for drinks

[lazy founder @FounderLazy]

@pHequals7 lol wtf it's way too early for that. i don't care the model is open sourced if it's shit. i can release an open sourced model by just finetuning some chinese model (which alex's team did previously with muse spark), it doesn't mean i'm worth $14b...

[lazy founder @FounderLazy]

@paulg i think important question is how did they reverse the trajectory

[lazy founder @FounderLazy]

@piyush100x you need to understand that not every attention is good for your startup. it might be when you're running B2C, but for B2B definitely not. so i suggest pivoting to B2C before ragebaiting

[lazy founder @FounderLazy]

@nadzi_mouad this guy predicted 16 of the last 3 releases

[lazy founder @FounderLazy]

smart people only need a well defined problem to solve. this is the 2nd most frequent way the most innovation happens. the most frequent by far is still pure luck (pre-conditioned by intelligence and perseverance)

[lazy founder @FounderLazy]

@mathemagic1an why not start with a system which does this in house, to isolate noise from the traffic, neighbors, loud music, etc.?

[lazy founder @FounderLazy]

@yacineMTB yeah, but it also says "high quality" data, so...

[lazy founder @FounderLazy]

fable still didn't solve thinking from first principles. it is like a junior employee with infinite knowledge - it can do whatever you tell it to do, but not really be creative

[lazy founder @FounderLazy]

they're not going to b2b with this since there is no ZDR for fable, unlike other models. i genuinely don't know what's the market for this - too expensive for individuals, too non-compliant for companies. i hope they just end up keeping it in the sub and this was the marketing trick

[lazy founder @FounderLazy]

@itsandrewgao yes on claude cowork very common

[lazy founder @FounderLazy]

@matt_slotnick @atelicinvest that would make sense if there wasn't a person with zero knowledge of ML leading their AI division. what's in fact going to happen is eternal catchup on benchmarks and no (even close) catchup on actual capabilities

[lazy founder @FounderLazy]

let's say you're a user for 5 years. spotify has in this timespan more than doubled their user count from 350 million to 750 million. just building a system which can take on 400 million extra users is a huge effort, not to mention fighting constant changes in dependencies, security issues (which exploded recently due to AI), etc.

[lazy founder @FounderLazy]

@yiguyuan20 @ChujieZheng @GanjinZero no he wasn't, he was just an IC researcher. "head of" means he had a team reporting to him

[lazy founder @FounderLazy]

@mil000 how can IG following be relevant for their B2B pivot? if they stayed B2C then yes

[lazy founder @FounderLazy]

Perplexity has built their own models as well (Sonar family optimized for search). I think they were onto something, all model providers are very shallow when doing web search and hence hallucinate a lot even with search, Perplexity could've positioned itself as the most reliable LLM-based search. Unfortunately, their infra is abysmal for anyone who tries to use their API, and that's where a lot of the money could've come from (at least initially - it's very hard to compete with Google and OpenAI for consumer market)

[lazy founder @FounderLazy]

What i'm trying to say is that principle is the same - startups don't work on fundamentals (revenue, retention), but rather on hacking the numbers to get the next round of funding. Pitching to investors becomes the skill in and of itself, same as writing grants. That said, since VCs are much more efficient capital allocators and, more importantly, much more hands-on in growing the distribution of startups than European accelerators are, their companies on average have higher success. But I think both are incentivizing the wrong skillset - the pitching skillset. This hurts a lot of companies that actually produce value (but struggle with "selling the dream" or "writing grants") on both sides

[lazy founder @FounderLazy]

we need children accounts on chatgpt like on netflix, my kid is polluting my memory @sama