6,552 karma · joined September 24, 2010
How many paired training images / text and what was the source of your training data? Just curious to know how much fine tuning was needed to get the results and what the breadth / scope of the images were in terms of original sources to train on to get sufficient musical diversity.
lynx -useragent="Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)" -head -dump http://www.example.com/
Not sucking isn't enough. Even being "Great" isn't enough. And even being "amazing" just isn't good enough to be successful. You need people to notice and care and help you and carry you to success. Mediocrity with a lot of public and positive reinforcement beats quiet greatness. Truly great things die every day from lack of attention and the willingness / motivation of others to help you carry that greatness forward. No greatness succeeds on its own, despite how great it might truly be.
OKRs are inane management fad. OKRs are "hard" because Objectives are vague and Key Results are vague leading to vague definitions of success and arbitrary flag planting.
Any management concept that is simultaneously so hard that no one does it right, but yet everyone is doing it is proof in point how valueless it truly is. Of course, when compensation, bonuses, and kudos are tied to OKR, people will keep doing them, as hard or as vague as they might be.
"Good OKRs tell a story about what is important, they help you inspire people to think about why they are working on what they’re working on, and I think they carry a bit of energy that comes from seeing how it all fits together in a succinct form."
That sounds more like a mission statement to me.
"Update: CNET's Caroline McCarthy tweets at us: "Sounds like a good theory, except that a Color exec told me Sequoia's $$ was nearly triple Bain's." So we're back to: HOW DID THIS HAPPEN?"
So Sequoia put in nearly triple what Bain put in. The article was just guessing before the update.
Some crazy stuff in this article by an Ag producer: https://agmoos.com/2021/12/10/supply-and-demand-are-the-real...
And competition between cheese, cream, ice cream makers for tight milk availability: https://nypost.com/2022/10/03/heres-why-the-us-is-facing-a-b...
When a bootstrapper tries to come into the market and isn't playing the investor-led startup game, by trying to build a company through customer money vs. investor money, they see that there aren't nearly as many customers as they would need or expect, and when customers are paying, the hype-driven, investor-backed startups are eating up a significant portion of the customer base.
This is because customers, far and few between in those rapidly growing, hypey markets, tend to be "lazy" buyers and will not spend the time to do any sort of real evaluation or analysis of competition and markets. Rather, if they have a real need for the product, they will buy either from who they are already buying from, or if they will select a new vendor, it will be someone they can get support internally to purchase (if it's an enterprise), with such support being a combination of a who their existing suppliers or consulting firms recommend, influence from content marketing or shilled analyst reports (buying influence), whatever companies are gaining the most attention and marketing at that moment, and/or who their colleagues and peers are recommending, all of which is subject again to startup influence, and besides a bit of a self-perpetuating cycle since the more they get recommended, the more they get recommended. In any case, it’s never about who has the better product or service or who offers the most value for the money.
In addition, and ideally, if the product costs nothing or very little to acquire, that makes it even easier for those few customers to gain support for a new vendor and that perpetuates the cycle, or if the customers' existing suppliers include the product or features of that product for no additional cost in their existing or new offerings, that makes it harder for the vendor. In any case, the fact that others are giving away a product for free or low cost makes it near impossible to compete as a bootstrapper, since you can't afford to give a product away, even if the startup can.
If you're selling directly to consumers (or to small businesses, which act like consumers) and not to enterprises, then it's even harder as a bootstrapper, since customer acquisition cost and complexity is very difficult and expensive, and those who have the most marketing dollars combined with products they give away or practically give away will win. Bootstrappers almost always lose in those market conditions. Yes, there are definitely exceptions to all the above, but I am finding the rule generally holds.
All this makes bootstrapping particularly difficult in fast-growth, hypey markets, which is where startups tend to proliferate. Basically, bootstrapping in an investor-led startup's game is not a recipe for success, and probably a road to ruin. You might be able to bootstrap selling ancillary stuff on the sidelines of what's happening in the hype-driven market, mostly marketing content-driven stuff, events, webinars, selling traffic, training or education, but even that stuff will come and go, and we're not talking large exits either. You can also sell consulting or advisory services in those hype-driven markets, but you have to really enjoy the consulting or advisory game and you have to mostly have relationships and connections to sell that sort of thing. Ideally, as a bootstrapper, you can go from hype-cycle to hype-cycle and keep and grow your existing customers. You can also bootstrap real products (not services or marketing content) in more sedate markets, but you'll most definitely face entrenchment of incumbents who protect their customer base, and so making entry into the market as a bootstrapper just as difficult. The key is to know what game you're playing and the best way to win at it, if you want to succeed, especially without exhausting yourself in the process.
I know someone who is actually following a different "third way". Rather than raising a few small rounds with aims of profitability (no investor will be satisfied with that since investors are looking for asset value increase, not profitability), it's about raising some early rounds, growing and getting attention fast, and quick exiting. He has been able to crack this code by building small, fast-growing, investor-led startups, but only up to a certain level (usually up to Series B), and then finding a quick acquisition exit, before the music stops. Somehow he has figured out not only how to do the quick fundraise, growth, and attention building, but also how to develop relationships on the acquisition end of the exit. While these aren't billion-dollar exits, this approach seems to work and he has a stable of investors who know how to play that game for everyone's benefit. I haven't figured that out myself.
Shows all the factors at play here:
* Company founded by kids who know / knew nothing about the space
* Investors who smell money opportunity and see an opportunity for asset value inflation
* Invest in the company, rush to multiple rounds, based on hype
* Company goes "public" using scam/sham SPAC approach (yet more proof of "startup" as financial product)
* Then the company goes bust after it goes to market because the whole thing was just asset inflation, and when the bubble pops, it goes poof
- I also find it curious that Techcrunch calls these obvious startup-as-financial product companies Upstarts and not Startups. A distinction without a difference in many cases.