He didn't clean it up when starting in the role, he damaged the (employer) brand he was expected to protect/develop as part of his role, he got re-assigned to a different role, which is arguably a reasonable consequence.
I'd say it is misleading rather than disappointing. The word is not capitalized in the Wikipedia article and titlecase is not a requirement here either.
It's not a requirement but it is the common convention.
Wikipedia has stricter style guidelines that go against that convention but in the absence of such a style guide on HN, one can see why common convention would typically be followed.
Agree in general though; I don't know why this convention exists and suspect we may be better off without it.
Yes. The budget starts at 1000EUR vs 250EUR. The difference is much less once you realize that you probably want two (disc?) breaks, probably suspension, larger wheel size. When I was making the purchase it was basically same price at that point.
Yes. It takes a few hours to learn to ride an EUC. But I'd wager that EUCs are safer because of that. Most accidents on rental scooters happen to first-time/occasional users. The fact that you need to learn how to ride an EUC gives you an appreciation of how bad things can go if you push farther than your skill allows.
In the end, EUCs are safer (larger wheel doesn't get locked in small potholes), more comfortable even without suspension, break faster due to more powerful motor, and are more compact.
Probably depends on the scooter. The segway ones I've ridden sound like plasticky junk and feel like it. The original bird scooters, on the other hand, are luxurious. They are xiaomi 365s. No suspension but it didn't bother me on this scooter with the ample tires. The brakes were great. If I'm able to lock the wheels while braking, that means the brakes are stronger than the grip of rubber to road, and the only way to get better braking is stickier rubber or an ABS system. A bigger motor wouldn't help here since the brakes already overpower the tires.
I would say durable rather than luxurious. The breaks were, but I'd say the tires weren't. I am yet to experience something like this on my v8f. There are other advantages of larger motors, like ability to ride uphill at larger inclines. Scooters are easily overpowered there and essentially useless in more hilly places. Bottom line is, that I am yet to meet a person who tried an EUC and prefers a scooter.
This is where people fundamentally misunderstand what hedge funds do. It's been nearly impossible to lose money as a long only. But hedge funds, which have massively underperformed as an industry have largely done so because the Fed has forced everyone to be maxi long risk assets. There is no alternative. And fundamentally, this isn't what hedge funds do.
I've often heard people say hedgies are just levered beta. Well, that's obviously not true as we've seen indices outperform. Some of the best long/short investors have been absolutely wrecked in the post 2008 QE environment (QEternity) because price discovery is meaningless. The Fed is in the driver's seat, and the managers who have killed it had one simple thesis: the Fed has your back, so BTFD. Front running the Fed has been the best and easiest strategy for over a decade now. And it will continue to be so until it isn't...
Not the GP, but I think they mean that there has been no point in hedging (hence "hedge fund") risk in the world of Fed-driven asset prices. If the Fed has your back to BTFD every time (and just look at the S&P 500 since the Covid crash -- a 30% loss was made up for in just six months), why bother trying to seek alpha to beat the market by trying to buy under-appreciated assets and sell over-appreciated ones? Just buy everything. And if you still want alpha, you just apply leverage. This last part leads to big volatility whenever there is a trigger.
Alpha should be risk adjusted (a la Sharpe/Sortino), but broadly your answer is correct. Current Fed policies punish those who don’t take risk, by design. Hedge funds (in theory - ahem) are structured to maximize risk adjusted profits. But if there’s no risk, because the Fed has made it so, then there is nothing for hedge funds to do but accept the new casino world for what it is: play along and hope that the Fed can control whatever comes our way. It’s a fools errand to hedge...until it isn’t.
Maxi long = maximum ownership that a relevant risk grid would allow
Risk assets = anything but the dollar or bonds
Basically buy assets which (they believe) stimulate the economy more by either lowering financing costs (“the hunt for yield”) or through wealth effects (“my portfolio is up 50%, I’ll spend/consume more now”)
> Ray Dalio’s Bridgewater Associates and Jim Simon’s Renaissance Technologies ... each suffered negative double-digit returns in their main funds in 2020. Similarly, some of the industry’s most highly regarded names such as AQR, Two Sigma, and Dimensional Fund Advisors all suffered massive investor redemptions in the face of substantial underperformance in 2020.
OneNote is not funcional. You can quite easily set it up so that your notes sync as PDFs to the cloud (Dropbox/Google Drrive/One Drive) via a3rd party syncing app. I use that and it works great. My notes sync to my desktop and I can add/remove papers from my desktop as well.
They probably mean that you can search in n-dimensional cubes, but what they'd like is having a more general shapes of the grid, e.g., have restriction of the sort: a+b<1.
Okay that is interesting. You could realize that by making a restricted area in the search space by returning np.nan in the objective function for those cases. Gradient-Free-Optimizers can handle np.nan and np.inf just fine.
This is at a high level how one of our Google internal black box optimizers behaves. The term used is "infeasible region" but it's the same idea as using a nan.
I would caution against using nan to always mean infeasible. Instead users should catch experiments outside the feasibility region and return a special infeasible value. This will increase visibility into the behavior of the optimizer, because it leaves nan to be used for values inside the region of constraint that are still problematic (due to bugs, numerical instability, etc)
If the volume of feasible space is small, the optimizer may have considerable trouble finding the feasible region(s) in the domain. The simple solution is to use a penalty function instead, i.e. add a term like M*max(0, 1-(a+b)) to the objective, with sufficiently large M. If the optimum is at the boundary of the feasible region, you might still get a solution that is slightly infeasible, which can be worked around with more special case logic...
Is there a subset of these constrained optimisation techniques that work when the constraint bounds are unknown before evaluation time? This applies to certain physical engineering problems where, due to the complexity of the environmental system you're modelling, you don't necessarily know what combination of input parameters lead to a non-viable solution. That is, a given solution might be numerically valid but the objective function could determine it to be non-viable due to not satisfying engineering constraints (e.g. it vibrates too fast).
Just adding a penalty to the objective works. Another alternative is to consider a two-dimensional lexicographic objective (total constraint violation, actual objective), so that reducing constraint violations is always preferable, and the actual objective is only compared between solutions that are equally good in terms of constraint violations.
This would be like the barrier function [1] for gradient based methods; which produces some challenges for those techniques. Do you know how this affects gradient free techniques?
Another option in some cases is to just map the space to something else. Like in this case, x and y are unconstrained and are what the optimizer sees, a = x and b = (1 - a) - abs(y).
Not always easy to do, but can work in some cases if the optimizer cant deal with it natively.
