avatar
B (🤿, 🤿)
1,476 Twitter followers
Follow
收益農民 explore financial primitives @ResearchNothing
Posts
avatar
B (🤿, 🤿)
Several points worth noting: 1) Once you experience the benefits of intelligence, it's addictive. OpenAI's free users use GPT an average of 7 times per day. However, when they start using the first subscription tier, the average daily usage doubles to 15 times; with the second tier, it triples to 21 times; and with the highest tier, it increases 11 times. This is the terrifying and addictive nature of the business models of these large-scale AI companies. 2) Computing power supply is still far from sufficient. "If you do not have compute, you do not have revenue." In 2026, OpenAI even proactively rejected some business opportunities due to insufficient computing power. Codex grew from 100,000 developers to 2 million in three months, with demand growing far faster than computing power supply. Only those on the front lines can truly appreciate how much you can do once you have computing power. For ordinary users, you can pay directly with subscriptions or APIs, but for enterprises and developers truly using AI at scale, in the long run, purchasing computing power is definitely a more cost-effective business than paying by token. The booming inference market has significantly extended the economic lifecycle of all computing hardware, increasing returns, and this process will continue. 3) Creativity and initiative are always scarce. Interestingly, however, OpenAI, an AI company, prioritizes people above all other resources. They are aggressively recruiting top researchers because computing power can be bought, and data can be accumulated, but people who can define the next generation of product direction and make groundbreaking breakthroughs cannot be bought. This proves that no matter how powerful AI becomes, its creativity and initiative still come from people.
Ark Invest Tracker
@ArkkDaily
04-01
OPENAI'S CFO SAYS: NO COMPUTE. NO REVENUE. - OpenAI is turning down business in 2026 because they don't have enough compute - Codex went from 100K to 2M developers in 3 months. - "If you do not have compute, you do not have revenue. That is one thing I know for sure."
GPT
0%
avatar
B (🤿, 🤿)
03-31
When everyone thinks Crypto is dead, will you still hold on? @darkmarketio's post-conference reflection on the Digital Asset Summit is very realistic and represents a common mindset in the Western Crypto community: DeFi founders are wearing suits, and protocols without traditional partners are finding it increasingly difficult to survive. Innovation is decreasing; most projects either repackage traditional yields onto the blockchain or chase trends, making money but rarely achieving VC-level returns. Many funds are struggling to raise capital, and a wave of liquidation may occur in the coming months. I can easily understand this sentiment. The situation is similar in the East, even a year or two earlier. To be honest, I've wavered, but Crypto itself hasn't changed; what has changed is people's attitudes. From initial ignorance and awe, to the belief in its omnipotence, inflated desires, and now being brought back to reality, we're starting to re-examine what it truly is. It's true that AI is draining liquidity and attention, but it's not just crypto that's being siphoned off; every industry is being affected. All the money and resources are flowing in. But just like when crypto initially swept the globe, every new revolution's capabilities will eventually have limits. Some say "Crypto is fintech now" has become the consensus, a downgraded narrative. But isn't it inherently finance? Looking back to when I joined this industry, it was simple: I simply saw it as a new form of finance—a more efficient, more open, and more promising version. Today, institutions are in, regulators have given it approval, and banks and financial institutions are embracing it, yet everyone is unhappy. In fact, people have had too many unrealistic expectations of crypto, thinking it will change the world. Of course it will, it definitely will, but in my view, it will be more on the right side of the balance sheet (the capital side). When global GDP (the left side of the balance sheet) starts to go on-chain, finance will naturally follow. This is a long road, long enough to dedicate a lifetime to. Till that day, Job is not done. twitter.com/bonnazhu/status/20...
ETH
0.51%
avatar
B (🤿, 🤿)
02-15
The screen is filled with interviews with Chase, the listing manager on ex-Binance. In an industry where long-termism is almost a joke, the equation "price = liquidity × attention × chip structure" is basically a slap in the face to all teams focused on product-market fit (PMF), revenue, and cash flow. However, most people in the industry understand these principles to some extent. Many are not surprised by the principles themselves, but rather by the fact that they are coming from a platform like Binance: "Ah, so even people at the top platforms in the industry don't think about the long term." However, there's really no need to overhype CEX's ability to select assets. Listing on CEXs often involves being swept along by the tide, unable to control one's own destiny, being pushed along by the market. @cz_binance made one correct point: Exchanges are essentially neutral; they are merely trading infrastructure that provides users with various assets, especially popular ones. This is perfectly natural, especially since the business model of exchanges is to earn transaction fees. Listing popular assets is also in line with the essence of the exchange business. However, at the same time, exchanges also bear a certain responsibility to "protect users," especially in this industry lacking regulation. Therefore, they have to conduct due diligence on teams and projects, trying their best to screen out teams that do not engage in malicious activities. However, the reality is that some projects are known by exchanges to be only short-term hype, but ultimately they are forced to be listed under strong market demand, user trading needs, and pressure to replenish assets. Wasn't that what happened with Meme? Having spent a long time in the cryptocurrency listing group, and having a background in research with an aesthetic appreciation for assets, I can understand all too well the conflict between long-termism and short-term interests. I have also seen the very realistic mentality within some exchanges: since most assets will eventually go to zero, it is better to list assets that can generate a wave of popularity in the short term, so that some people can at least make a profit, which can also be considered as giving users and the market an explanation. It's understandable that many people have expectations of Binance. After all, it's the industry leader, the largest liquidity portal, and plays a pricing role. With great power comes great responsibility, so naturally, people hope that in addition to making money, it will also provide some kind of directional guidance. But in reality, Binance alone cannot solve the problem: 1) In traditional finance, asset screening is not the responsibility of exchanges, but rather investment banks, securities regulators, and a host of laws and regulations. However, in the crypto space, these entities have long been absent. 2) From the perspective of the nature of assets, regardless of whether they are crypto or not, all assets are inevitably affected by liquidity, attention, and asset structure, to varying degrees. Just look at the erratic behavior of speculative stocks in China's A-shares market; this isn't unique to crypto. It's just that most crypto assets are in the very early stages of their business lifecycle, anchored either to startups or the narrative itself. Therefore, "going to zero" is the norm. There's no need to elevate it to the "original sin of crypto or Altcoin." What did Altcoin do to you? Stock exchanges are not venture capitalists, much less gods. It cannot guarantee a long-term winning rate for any asset. Moreover, is VC's win rate high? It's pitifully low. What I want to say: Crypto is simply one form of asset. If it's primarily anchored to emotions, narratives, or imagination, then the volatility of liquidity and sentiment will naturally be amplified. If, in the future, it gradually anchors itself to equity or the cash flow of a growing business, then its pricing logic will change accordingly. The real questions we need to answer are: Should a project issue crypto? Why might issuing crypto be better than traditional equity in the next era? And how can we attract more mature, higher-survival-rate businesses, companies, and assets to the crypto world? The job is not done! get back to work
avatar
B (🤿, 🤿)
01-22
Taking advantage of a weak market, act decisively. Shifting from the Ve token lock-up model to a staking model, thus facilitating a quick update of the token holder structure at a low point, is a good thing; there are fewer convoluted processes. 1⃣ Is the Ve model the only solution for governance? Yesterday, Michael, the founder of @CurveFinance, was saying that abandoning the Ve model was a mistake. Of course, the Ve model's "hard lock-up + voting" is indeed a model: if you want to influence pool listing and pool emission for your own benefit, you have to pay sunk costs—this logic is sound. However, it's not without weaknesses. If you carefully examine the token holder structure of most Ve models, you'll see that most involve external project teams + third-party voting bribery platforms, because they need to drive emission to increase the TVL (Total Value Limit) of their pools. Therefore, this system was initially designed to "harvest" these big players. Moreover, most of these big players ultimately use third-party voting bribery platforms to allocate budgets, allowing platforms that have locked up governance tokens to vote on their behalf. It would be better to be more direct and allow projects to directly incentivize their own pools. Uniswap, for example, had its own LP Programs in its early days. As for the remaining emissions, besides saving some, they could be directly handed over to the market instead of being decided by voting. Higher TVL and better trading volume mean more, and vice versa – that's fair. Ultimately, the Ve model itself has inherent costs. Even for most traditional investors, the idea that "only by locking up money for several years can one participate in governance or receive dividends" is a very counterintuitive design. For Altcoin to break out of its niche, it needs to expand its holder base like Bitcoin. Simply focusing on the advantages of Ve and only allowing a small group of users to participate creates a stabilizing structure. 2⃣ Hard Lock-up + Voting + Bribe > Buyback + Soft Staking? There's no standard answer to this question. During the transition period, there might indeed be some pressure from the unlocking of hard lock-up tokens. However, Pendle's VE staking ratio isn't very high, unlike projects like @CurveFinance or @AerodromeFi which have exceptionally high VE staking ratios. In reality, many staking projects have much higher staking ratios. Therefore, in equilibrium, the former (Ve) isn't necessarily superior to the latter (Ve). More importantly, supply and demand are not something that token models, as mere "techniques," should excessively bear. Relying solely on token design to "adjust supply and demand" is almost always unsustainable. What truly determines supply and demand is always the product itself, revenue expectations, and liquidity. Token models cannot solve these issues; they can only amplify or distort them. From a purely pragmatic perspective: The Liquid staking model is simply more comfortable than the Ve model. At least, that's how it is for me. twitter.com/bonnazhu/status/20...
PENDLE
4.56%
loading indicator
Loading..