Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×

Comment Re:cost of doing business? (Score 1) 67

Examining the description posted by the SEC, it does look pretty accidental

7. In June 2019, a team at JPMorgan within the Corporate Compliance
Technology department (“the eComm Tech team”) was working on a project to delete old
electronic communications, including emails, instant messages and communications conveyed
over Bloomberg that were no longer required to be retained. Using policies and procedures
developed by JPMorgan and technicians at the vendor to facilitate the deletion of documents,
the eComm Tech team tried unsuccessfully to delete certain historical communications from
the 1970s and 1980s.
8. To troubleshoot the process, the eComm Tech team spread out deletion tasks
across multiple time periods and ran deletion tasks on emails from January 1 through April 23,
2018, erroneously believing, based on representations from the vendor, that the 2018 electronic
communications required to be maintained were protected from deletion by the vendor’s default
retention setting of thirty-six months. In fact, however, the vendor had failed to properly apply
the default thirty-six month retention setting to the “Chase” communications domain. As a result,
the troubleshooting exercise permanently deleted all of the emails in that domain from that
period which were not subject to legal holds.

JP Morgan paid about %175 million for a more malicious case of using WhatsApp a few years ago

Comment Re:"This one guy" caused the stock boom? (Score 1, Insightful) 31

...This is not some fantasy and contrary to your claims it's a tactic used by short sellers to increase profit. They do suggest in this article it's also to protect investors but I suspect altruistic short selling is virtually non-existent in reality.

I mostly disagree with your assessment, but this sentence is gold.

...Then there's naked short selling where you're shorting more shares an actually exist which hugely distorts the price as without true supply demand price discovery doesn't exist.

In the electronic era, a little bit of naked shorting does still exist, partially due to code bugs similar to the one that cost Goldman a $3MM fine this spring. But naked shorts are easy for regulators to detect. Naked shorting is just not in volumes significant enough to affect market price. At least for companies worth writing about -- i.e. with nontrivial market capitalization, outside the "pink sheets" / penny stocks which are the apparent focus of cellar boxing (which I had not heard of, because I have never dealt with companies whose price is close to 0.0001).

Comment Re:Medical dogma (Score 2) 17

And then as we saw during the pandemic, there were plenty of doctors who refused mask mandates, and provided false vaccine information (ivermectin and HCQ were not just spread through social media - many doctors were guilty of it as well).

Those doctors were definitely a problem, though I would argue generally a different problem than medical dogma. Getting rid of foolish conclusions arising from motivated reasoning is...really hard. I'm more optimistic about mitigating dogma.

Comment Medical dogma (Score 4, Interesting) 17

I get the sense there is a tremendous amount of medical dogma that gets passed on even to today's MDs. The most famous example of course was the cause of ulcers (dogmatically due to stress, found in the late 20th century to be bacterial). But a recent famous example was the medical community's deadly dogma concerning aerial viral transmission, brilliantly covered in this Wired article and subsequently broadly discussed elsewhere. This example of tumors containing bacteria has the same flavor of scientific mispractice.

There are a lot of beliefs in the medical community resting on shaky old studies like that, or on no research at all. The typical medical education consists of tremendous feats of rote memorization rather than scientific thought (just ask any MD how they got through organic chemistry).

I generalize here; I know this to be a problem for US medical doctors. I am not sure to what extent the rest of the world recapitulates the issue. But since France for example is full of actual doctors recommending homeopathy(!) I am not optimistic.

Comment Re:Fine whisky? (Score 1) 135

I'm with you. Stuffing as much peat smoke into a whisky is like cramming an IPA with too much hops. It is very easy to overdo, and some people just want to show off. I'm a fan of Black Bush for a good Irish whiskey or Glenkinchie 12-year if going for scotch whisky. I find lowland scotches much more palatable than highland or Islay scotches.

Comment Re:And nothing will happen (Score 3, Insightful) 79

Google will get a minor slap on the wrist, say I am sorry, and admit no fault.

You're probably right, but sometimes these things get serious. Last year, US investment banks ended up getting fined a collective $1.8B for not having copies of employee messages (including $200M for JP Morgan and Morgan Stanley).

That was the SEC, though, and regulatory rather than legal. So it's hard to foretell how Google will be treated.

Comment A high failure rate means a higher failure rate (Score 1) 123

As far as Apple is concerned, the rejects never existed. If a local company wants to shrug it’s shoulders and accept a 50% reject rate, they just jugaaded themselves into half-the-revenue-they-could-have-made.

No harm no foul for anyone. Apple gets what it wants.

It's not quite like that, because Apple's own quality control is necessarily imperfect. Let's say that the India manufacturer constructs 2,000,000 cases, of which 1,000,000 pass internal QC and go on to Apple, and that the manufacturer's internal QC has a 5% miss rate. Then Apple got 50,000 bad cases. If Apple's own QC checks have a miss rate of 1%, then 500 bad phones go out to end users.

In contrast, if the Chinese manufacturer constructs 10,000 bad cases and 990,000 good ones, sending them all to Apple even without having any internal QC, then only 100 bad phones ultimately go to end users.

Comment Explicit policy (Score 1) 108

I teach as an adjunct in a masters' program at UChicago. Just this quarter, I added an explicit A.I. policy:

Since your homework and projects require both computer code and prose, you may find A.I. tools such as ChatGPT helpful. We encourage you to try them out. If you employ A.I. to help with your homework, please indicate the prompts involved.

I will be interested to see what effect this has,

Comment You mean "hard statistics"? (Score 1) 80

"Since 2015, *right-wing extremists have been involved in 267 plots or attacks and 91 fatalities*, the data shows. At the same time, *attacks and plots ascribed to far-left views accounted for 66 incidents leading to 19 deaths.*"

It has only gotten *worse* since then with the right ramping up the violence across the nation dramatically in the last 2 years.

https://www.washingtonpost.com...

Slashdot Top Deals

After a number of decimal places, nobody gives a damn.

Working...