I like how the article did its best to bury the lede that the crux of the discussion was how to better understand and manage these new technologies. To quote the panel prompt "[T]here are some real-world fears and ethical questions that need to be asked. Just because we can, should we?"
Isn't this exactly the type of discussion we want happening in a public forum?
There's a perverse incentive here where the military industrial complex is constantly pursuing more profits/contracts. That almost always means inventing new, terrible weapons. So the ethics end up taking a back seat to corporate greed.
The recent curative hepatitis treatments have paved the way for this. There are a variety of different approaches including statewide cost sharing funds that pool the cost across all the insurers in the state.
IIRC, single-payer health care systems like the NHS do functionally the exact same thing, which allows them to negotiate significant discounts. I remember reading about a gene therapy for a type of leukodystropphy that was in the millions of dollars, but got negotiated down. [1]
Paperclips and Crude Oil value, relative to another paperclip or barrel of oil, can be based on well-defined, finite, measurable standards that can be measured indeterminate of their location. That allows pricing to be indexed based on location, time, and quality relative to the standards.
A house's value relative to another house is an almost infinite number of variations of measurable and immeasurable variables. One example-- Traffic noise. Maybe the amount of noise is an issue for you, maybe the time of day of noise is an issue for me, maybe noise in one room vs. another. Moreover, there are very expensive switching costs for me
I would argue that the cost is less in setting up than in running and maintaining a site at the scale of Recreation.gov. That includes aspects of customer support.
Many of these agencies are forced into uncompetitive compensation structures which means contracting out most, if not all, of their technical work.
A major part of the issue is the government contracting system. In an attempt at fairness, it has massive amounts of oversight burden. That is, in turn, a barrier to additional competition for the work.
Because they aren't writing for a general audience. They're writing for an academic, professional finance, and governmental audience. It's extremely relevant that their word choice is being scrutinized for suggestions of what their future policy will be. Similarly, their use of jargon would be well understood by their audience. Context matters in communication.
You could consider this almost ritualized. That's something very different than corporate doublespeak that seeks to obscure rather than inform.
As with most fields a lot of it relates back to the first principles of the field. A lot of it is around what the best way to classify, measure, and show an economic transaction. One example that comes to mind is the development of Activity Based Costing in the late '80s as an improvement on traditional cost accounting.
You'll see similar models in law firms. IMHO it's driven by two key elements in the business: An apprenticeship-like, up-or-out model for developing talent and a vested owner-operator class at the "partner" level.
These firms effectively operate as a collective for the benefit of the partner class. New partners must invest capital, ongoing partners get returns, and older partners retire with a terminal payout. Some partners may take on expanded leadership roles and get higher annual payouts.
The overarching legal structure is a sort of "modularization" to limit the liability of lawsuits and enable flexible operation. There is a usually a global partnership council that dictates universal operating norms and practices.
Again, fundamentally, it operates as a collective and international units support each other for collective benefit.
Can you elaborate on the "new partners must invest capital" part? I've heard of this, notably in the TV show Suits where when the lead character got promoted to partner, he was asked to contribute a lump sum of cash. Sounded nuts to me that a promotion meant you basically got a chance to pay (I know it's actually invest) money.
Every firm has its own specific way of doing it, but the short version is that non-partner workers at a firm are normal employees with a normal salary, benefits, etc. But they do not own any equity/stake/shares in the firm.
Once you are promoted to partner, you are no longer an employee of the firm and in some cases you no longer earn a salary. Instead when you are promoted to partner, you are allowed to "buy in" by purchasing a certain number of shares of the company. Once you own those shares you are then part-owner of the firm (rather than an employee) and enjoy profit-sharing.
Relative is in HR at a big4, this is basically correct. Some ballpark numbers.
A "first year" partner may have to pay ~$150,000 into the partnership, on perhaps a $300,000ish a year in profit sharing. Usually this is just a reduction in the profit sharing (or perhaps a multi year loan, i forget). Its not uncommon to earn less as a first year partner than your last year as Managing Director.
You are not longer an employee, but part owner. You move onto a totally separate HR system (Benefits, etc.).
Relative works with mostly senior partners, multiple of them are making $1 million+ a year. That's not common, but as you move up the ranks of partnership, that's not a crazy salary. They work a lot, are all divorced, don't seem to even have time to enjoy the $$ etc. Listening to the details, I'm not super jealous.
It may be different in other firms but at least for the big 4 firm in familiar with sr managers can be promoted to either partner or managing director, and mds usually take home more that first year because of the buy in, usually in the form of an interest free loan.
you have to buy a partnership. The firm usually makes you an interest free loan to buy the partnership. Some people bought into arther anderson right before enron blew them up.
The partnership is often the most valuable single asset a partner owns. It prevents them from jumping ship and taking clients. And it acts to hopefully prevent a partner from taking risks that could destroy the firm. Even senior people who might become partners in the near future will not want to put that partnership at risk.
When you retire, they'll buy your partnership out - often paid out over a few years.
That doesn't feel much different from the (common, highly tax-favored) early exercise of stock options in a startup.
If you're becoming a part-owner, of course it makes sense you have to buy shares, my only confusion is how they determine the relationship between the value of those shares and what you have to pay for them.
They know how many shares in the partnership they have, and the profits they earned and have no other shareholders, the remaining profit gets divided up among the partners.
I would be really interested to see this model applied to software. I am unaware of any firm that has done it this way so far — which is odd, because it really does seem tailor-made to suit our sort of work, no?
i think the reason consultancies (legal firms, accounting firms, etc) do it, and not traditional software companies, is because the worker hours translate directly into billable hours. Partners are incentivized to bring in more business, which the associates (and the partner him/herself) can bill against. I think partners are usually expected to bring in a certain amount of business to justify their profit share.
In a traditional software company, like a SaaS business, I don't know how it would translate. A developer fixing bugs might be important for the business, but ultimately is hard to translate to the bottom line. It's not like hiring a high level executive will have a guaranteed ROI in the same way.
Sounds like the distinction is between project work and owning one or more products. A Wipro could function like this. No reason a Microsoft couldn’t, but they just don’t.
Deloitte make a huge point of hiring apprentices - taking on dozens (maybe nearly ~100) of them on every year. Seems the general goal of everyone there is to make partner one day.
In software co's, profit sharing is split between equity vs commission, where equity for hard-to-attribute & motivate stuff and commission for easy-to-attribute-and-motivate stuff. Doing profit sharing incentives for devs risks getting weird quickly.
-- It's hard to measure contributions of a dev in a team of 5-10 (how much was them, another dev, the PM, the designer?) while pretty clear for say a sales + sales eng pair.
-- Likewise, there's a danger of confusing incentivizes of having a dev make commission-based decisions, esp. when counter to what a PM needs to incentivize. in a sense, part of the PM's job is to prevent sales/marketing/management from confusing the rest of the org!
For product orgs, some companies do incentives and promotions based on profit/loss measurables ("1% datacenter power savings => $10M/yr savings => ..."), I think team-based w/ some sort of hierarchy. Alternatively, Cisco weird spin-in system worked well... for those liked by the involved leads.
When _not_ a product org and a more direct line to $, like a sales eng supporting sales, or a dev supporting a quant, easier to go commission/profit-sharing based..
I always wondered why law firms used partnership structures, but this seems like the core reason (most other reasons could would still work within a limited liability corporate structure).
I know opportunity costs are a thing but a lawfirm seems like the one of the worst possible targets for a lawsuit even without attorney-client priveledge complicating matters. Barring smoking gun damning evidence (which they would be skilled at knowing how to avoid both in terms of not doing it and not getting caught).
I would have guessed it as more a result of billable hour basis means it is practically "pure per person labor" dependent even with structures of management, less experienced lawyers and paralegals to effectively transmute the less experienced working hours into more experienced hours by them reviewing and signing off on the work of underlings.
Without the head partner you have 400 billable hours of junior lawyer and paralegal work. With one you effectively have 410 hours of partner legal work and they have staked their reputation and used their expertise on it to verify it. If the partner spent those 10 hours day drinking and signing off without reading it turns out to be 400 hours of soverign citizen tier pseudolegal nonsense arguments is their head essentially.
> I know opportunity costs are a thing but a lawfirm seems like the one of the worst possible targets for a lawsuit even without attorney-client priveledge complicating matters.
If you're suing someone else's lawyer, yah. But if you're suing your own lawyer... that privilege is client's privilege, and they can waive it if they want to sue their attorney (AFAIK).
there is a concept in RICO law called 'piercing the corporate veil,' where lawsuits against a company can 'pierce through' layers of holding company structures, to exactly prevent people from creating shell companies to limit liability.
In practice, however, it is another line of defense, because you have to argue in court for each layer to peel back.
In the U.S., law firms are very often structured as "limited liability partnerships," a.k.a. LLPs. In most versions of LLPs, each individual lawyer is responsible for her own malpractice and that of the people she supervises, but not of the malpractice of others. [0]
Do they ever clean house of one of the national divisions? Or wouldn't do that (too directly) because that could be construed as being too involved in the local operations?
Companies that file for bankruptcy don't have much damage done to operations and the organizational structure because its ability to keep operating is usually the highest value outcome. That is- the company's continued operations bring in more money than selling off the assets. Most corporate bankruptcy is about not making enough money to pay their outstanding debts, not having money losing operations.
Typically the owners lose all their investment and the debtors lose some of their investment. The debtors, in order of priority, are given ownership of the company by the court.
The ability for ownership to exist separately from the "operating" company, for the owners to only be liable up to losing their investment, not more, and for the management and employees not to directly lose money is sorta the "why" for corporations to exist.
Lots of countries don’t have restructuring codes. When their corporations go bankrupt, the assets automatically go to the creditors and/or are liquidated.
This material is concerning to me on a base level.
To me, it doesn't reflect positively on the cost and complexity to implement a large scale system. It presents much of the material without a simple guide to what should be optional, when, and where.
You almost need a huge number of people to even digest the material and then make decisions about it. It feels like a massive reflection of the government contracting industry. I see it illustrating the value of editorial or architectural control and a willingness to take risks by saying "no".
Systems engineering in its very nature is a huge task. You’re trying to balance the various tug of war battles between everything. Cost, performance, reliability, weight, etc... How light is this material vs how strong is it? What kind of tooling do we already have that can we use vs do we need to invent new tooling? Inventing new tooling is expensive. Inventing new tooling requires training on new tools. Training is expensive. The list goes on and on. Systems engineering is preventing deadlock.
I’m speaking in the physical world (aircraft) but the same concepts apply to software. If your system is small, sure you can do it all in your head. But if it’s large, you need to put a cohesive plan together that accounts for everyones issues, even if they aren’t going to get solved right away.
So yeah, you do need a whole team to do this kind of thing.
