这两年来，奇绩创业营一共收到了 36000 多份的申请，这些创新领域的数据都不断增长，我们挑了一些比较有说服力的数据来跟大家分享。

When web3 becomes a credible alternative to web2 for the masses, large centralized companies like Facebook, Apple, Amazon, and Google will have to compete for access to our data thus redefining how we use the web.

In each technology wave the value, revenue, market cap, profits and great people captured by startups versus incumbents differs. In some waves it all goes to startups, while in others it goes to incumbents or is split between them. Unexpectedly, the prior wave of value from AI roughly all went to incumbents over startups, despite a lot of startup activity.

关于网状安全性（Mesh security）模型设计，是什么可以阻止其中一条 Cosmos 链成为整个网络的核心，然后获得网络效应以及流动性溢价（liquidity premium），从而偏离菌丝网络的目标（例如，目前 Osmosis 就是整个 Cosmos 生态的流动性中心）。

DAO。 开源社区，有点类似于公司与 DAO 之间的中间过渡形态，是在加密货币还没有诞生之前，基于法币的一种「打造出更高生产力/满足更新需求的软件」的折衷方案。

As mentioned above, games such as World of Warcraft and Dream Western Journey have been popular for nearly 19 years, some MMORPG games in Roblox also have lasted around 10 years. Trading is at the essence of MMORPG games, emphasizing the concepts of assets and in-game economy.

MakerDAO、Aave、Compound 等龍頭項目從容不迫、按章辦事，無論是治理還是清算都井井有條；Solend 、Maple 等新興項目狀況頻出、問題不斷，暴露了自身短板，也揭露了所處領域的頑疾。我們也希望，這些經驗和教訓能給 DeFi 從業者提供參考，推動 DeFi 進步。

If there is a safety assumption of 51% on the N nodes of the oracle, then the overall safety assumption states that to compromise the system, the adversary must compromise (51% * N) && (1) — that is, the (one) “Relayer.” If the adversary can corrupt 51% of N, it might as well corrupt (51%+1) nodes, especially given it has the option to run its own +1. So the “1” node doing transaction relay doesn’t buy more economic security. Instead, it introduces poorer liveness characteristics, since the whole system loses liveness if one node is compromised. It would be better to change the 2/2 multisig into a 1/1 and to strengthen the oracle’s own security with an extra node. We know Proof of Stake can carry better security and liveness than a multisig (otherwise all these blockchains would be running on simple multisigs!)

I’m interested to see if the complexity around rotation of validators causes issues in the future, as the protocol needs to rotate private key shares and full nodes when a validator joins or leaves.

If there has been quite an active OTC market for the locked coins and ‘weak hands’ have sold to higher conviction investors, then an unlock event is effectively just the removal of “fear”. 

As demand increases for unlocked coins and the market cap increases, the FDV increases proportionally, even though demand for these locked coins does not necessarily increase. As a result, the FDV increases 1:1 with market cap even if those locked coins could perhaps have happy sellers at much lower prices.

Many will lose faith and won’t be able to stomach the soul crushing multi-year grind lower that is a crypto winter. “Wow, the government might actually regulate this out of existence,” “It’s just too early for these products,” and, of course, “I told you this was a bubble” will be among the drumbeat of negativity you can expect to hear parroted by critics. In addition to eating big paper (or real) losses, you’ll see people have breakdowns, go bankrupt due to overleverage (or poor tax planning), quit otherwise promising projects, turn nasty, depressed, or apathetic, and generally lose sight of the longer term potential of crypto. To make matters worse, the next bear market will be a regulatory nightmare, and we won’t have the bull market vibes to help defend ourselves against all of the consumer protection, fraud and abuse, systemic risk, ESG, and illicit activity FUD that our enemies will throw at us. At the same time, the “grassroots” crypto herd will thin because it’s tougher to wage war when you’ve lost 90% of your savings and need to go find a real job again.

By many metrics (price-to-sales and price-to-earnings), DeFi remains compelling, but the math only works for whales right now.

That said, inter-protocol competition is fierce, regulatory scrutiny is coming, technical vulnerabilities are pervasive, systemic defaults could cripple the entire market, and high gas fees are crippling the unit economics.

1) historically low-interest rates and (2) the winner-take-all nature of many technology-powered businesses.

these hospital cost challenges unfortunately converge with what we believe are increasing inflationary pressures across medtech, making it significantly harder for medtech to pass through those inflationary costs to their customers.

廣告在與 app 功能和用戶探索過程中的整合程度決定了廣告轉化率和 CPM (每千次展示費用)，我們認為 app 的「實用價值」變得比使用者在 app 上花費多少時間來得更加重要。Youtube 和 Tiktok 相比其高黏著度的用戶使用時間為什麼相對被「低估」，正是因為它們都是以最大化使用時間或觀看時間做優化和設計的 app，使其娛樂價值高於實用價值（而因此被低估）。Gamma 認為以使用時間來衡量社交/娛樂 app 的日子將會結束。

SushiDAO Lending Pool on Maple

As seen in the Sushiswap transactions above to swap SUSHI directly for ETH, a market sell would incur ~30% of slippage impact to the buyer and the overall market. For USDC, it would be ~43% slippage. The SUSHI in this market sell scenaro would be immediately liquid for the SUSHI buyer vs. the SUSHI sale proposed in this raise would be restricted for 2 years at a 20-30% discount.

The model takes into account SUSHI’s historical volatility, crypto market volatility more broadly, and the proposed vesting schedule to back into an adjusted value for SUSHI under those liquidity terms.

Unlike Ethereum's TVL, which has a healthy dose of stablecoins in the mix, the composition of BSC’s TVL was heavily skewed towards the higher end of the risk spectrum, making it extremely sensitive to market swings.

LAO 也採用了 Moloch 的框架，使該組織能夠接受資金，而不僅只有支付功能。到目前爲止，這個項目已經吸引了 2500 萬美元的資金；另外，它已經在 40 個項目中投入了大約 30% 的資金

During its IPO roadshow over the weekend, CFO Jason Warnick outlined a number of prospective products, including a new wallet, a lending product, debit rewards and staking.

the 600 basis points ofgross margin expansion is another great validation of the growing efficiencies that we'veachieved through product design and efficient manufacturing operations

And what do they think will serve the nation as a whole? My guess is: Power. Geopolitical and military power for the People’s Republic of China, relative to its rival nations.

It’s possible that the Chinese government has decided that the profits of companies like Alibaba and Tencent come more from rents than from actual value added — that they’re simply squatting on unproductive digital land, by exploiting first-mover advantage to capture strong network effects, or that the IP system is biased to favor these companies, or something like that.

But notice that China isn’t cracking down on all of its technology companies. Huawei, for example, still seems to enjoy the government’s full backing. The government is going hell-bent-for-leather to try to create a world-class domestic semiconductor industry, throwing huge amounts of money at even the most speculative startups. And it’s still spending heavily on A.I. It’s not technology that China is smashing — it’s the consumer-facing internet software companies that Americans tend to label “tech”.

China’s attack on its tech companies, in contrast, seems far more comprehensive — it’s not just attacking the biggest internet companies, it’s attacking the entire sector. (Update: An important piece of evidence here is that China also appears to be reducing venture funding. If you want more competition you don't squash new entrants!)

Netflix isn’t trying to increase its average revenueper subscriber by offering newsubscriptions plans; it wants to keep raising theprice of the subscription plan that it isalready offering.

Start with the IP point: this is themost obvious reason for Netflix to gointo games, but honestly, I find it one of the leastcompelling ones. In a vacuum the bestway to monetize film and TV IP in gaming is to licensethat IP to dedicated gamesdevelopers, which is exactly where Disney, after yearsof trying to build a gamesbusiness, has ended up. Building a great gaming experienceis a fundamentallydifferent skill and, for Disney, a distraction fromtheir core business

ER visits improved 1,390bps sequentially over 2019 (to -5.5% in Q2 versus -19.5% in Q1) suggesting a more normalized operating environment, and outpatient surgeries improved 740bps to levels above 2019 (to +3.4% in Q2 versus -4.0% in Q1)

While ZTS is facing particularly easy comps (its business saw a significant impact due to COVID last year) and restaurant demand appears to be rebounding, producers are also facing rising input costs (corn +37% YTD).

The company’s dual reference lab and point-of-care offering representcompetitive advantage relative to peers, and Idexx has significant runway to develop international markets over time.

animal health multiples at all-time highs, we believe upside to numbers (which we expect) will be needed to support current valuations and sustain the moves we have seen across the group.

strong pet market fundamentals (which are showing no signs of receding) and easy comps (as reminder, last year’s 2Q results were impacted by meat processing capacity constraints on the livestock side and vet clinic closures on the pet side), we are expecting another set of very strong results for our animal health names.

twasclear to us from both Mr. Sayerand Mr. Leach’sinterviewsthat “capability” is a key word in the phrase, as the feature “won’t go there on first pass” and will instead come to market through an app update later on

Bottom-lineis we understand why some investors may be getting frustrated with O5’s extended approval timeline, especially as the product’s pivotal trial data has looked so strong and we’re now 3-4+ months past what most investors had expected would be a mid-1H approval for thissystem. And while we also don’t disagree with investors whoargue that management has sounded fairlyconfident over the past few months about O5’s “imminent” approval, we believe it’s unfair that some of those investors are beginning to argue that management’s credibility is starting to suffer due to the heretofore lack of O5 approval.

Like many investors, we often question and lament why these kindsof app-related deals and integration efforts seem to take so long to execute and result in a marketable product. The bottom-line, though, is that they simply do take time (likely due, at least in part, to data privacy and HIPPA issues). Even so, it is clear that DXCM remains in the lead on the data integration and now API front in the CGM market, especially when compared to ABT, where integration with external apps remains limited and even the company’s own apps that are available in the U.S. (LibreLink, LibreLinkUp, etc.) have pretty low scores on the Apple and Google Play app stores.

The key swing factor going forward will be the timing of when reimbursement parity between digital and in-person visits will be reassessed and by how much. Our follow-up conversation with management suggests a decision hasn’t been made yet and that virtual care reimbursement will likely depend on the type of service and value created.

telehealth continues to be a strategic priority for UNH, who has delivered 500K virtual visits through its OptumHealth behavioral platform year-to-date. Virtual care encounters have had a patient satisfaction score of 98%, suggesting telehealth utilization is sustainable post-pandemic, offering accessibility and convenience.

the role that shifting sites of care and consumer preferences could have on utilization and MLR going forward. For example, home care was estimated to lower hospital admissions and post-acute medical costs by 25%, while drugs that cost $10-15K to administer in hospital outpatient care could be delivered by Optum Home Infusion for $3.5K.

With the increased use in telehealth, lower ER visits, and proliferation of primary care models that aim to lower hospital readmissions, we see a higher likelihood of healthcare utilization settling at a new baseline that’s lower than pre-pandemic levels.

Commercial core utilization peaked in April, declined slightly in May and June but remained marginally above baseline levels by the end of the quarter (negative read-through to CI, CVS, ANTM). Medicare and Medicaid tracked a few percentage points below baseline (positive read-through to HUM, ALHC, CNC).

Tools have also been topical, and have been recently favored -- with the group having led performance in HC over the past month, the bar is now seen as high for incremental gains on 2Q prints in-and-off themselves (see preview here), though we sense print-related dips could prove short-lived given ongoing healthy appetite for exposure to this group for both micro (intact normalized post-covid growth compounders) and macro (potential inflation beneficiary) reasons.

feedback following our mid-year medtech review/2Qpreview (link) suggested alignment with our thinking on the quarterly cadence from here (in-line 2Q prints, 3Q guides need to generally come down) but more varied in reaction to our 2022 optimism

persisting bear market in SMID biotech marked by a lack of generalist follow-through amidst high frustration from the dedicated/specialist Rx community who have been getting caught wrong-way. This has been particularly evident over last month, as book-ended by MRNA's ongoing march upward on one side (the new TSLA? Exhibit 5) vs. the ongoing unwind in BIIB on the other (now down ~25% since the early June Adu approval).

UNH (which hits these screens) following its 2Q earnings served somewhat instructive as a potential early clue suggesting that the quality-theme may continue to dominate print "noise"

our investor conversations continue to point to limited appetite to take significant "print-risk" with low conviction on what will dictate price action on T+1

A bid in large cap "quality" names has persisted on what seems a broader rotation to safety amidst ongoing debates around a nebulous macro (daily flip-flops on growth/value; covid on/off; inflation on/off, etc.) that could linger through Jackson Hole in late August

Overall Medicaid enrollment is up 23% from pre covid levels seen in March 2020. Public exchange benefitted from special enrollment periods and expanded eligibility and subsidies from the COVID bill, with enrollment up ~15%. We expect employer enrollment should be stable to modestly up among the major employer MCOs.

Profitability movements worth watching for minimally invasive BPH (Teleflex & Boston Scientific): As we lay out in Exhibit 1 below, proposed payment rates for CPT code 52441 (first UroLift implant) and 52442 (each additional implant) were down 24% and 26% to $1,155 and $796, respectively; similarly, CPT code 53854 for Boston Scientific's Rezum was down 23% to $1,528. Importantly, our preliminary profitability analysis (see Exhibit 2 below) points to a shift between the two devices where Rezum's profitability (~$375) comes out above Urolift, which has dropped by almost $1k to $100-125. That said, we see it as too early to call whether this could translate into a shift in physician utilization. Specifically:

不過房租項目中，走高主要來自於自有住房換算成租金的項目上揚，亦即反映房價的上漲，因此預期這在短期對於實際消費能力的負面影響將較為有限。

因為蛋白質的功能取決於其立體結構，而目前根據已知某基因序列可翻譯獲得對應蛋白質的胺基酸序列，即蛋白質的一級結構；如果從蛋白質的一級結構就能知道立體結構，那麼即可直接從基因推測其編碼蛋白質所對應的生物學功能。雖然蛋白質可在短時間中從一級結構摺疊至立體結構，研究者卻無法在短時間中從胺基酸序列計算出蛋白質結構，甚至無法得到準確的三維結構。因此，研究蛋白質摺疊的過程，可以說是破譯摺疊密碼的過程。