In the news

More specifically the dubious nature was picked up by analysis of the images in the papers. Many scientific papers use doctored images of results to justify their findings.

See the following article: https://www.nature.com/articles/d41586-020-01363-z

Exactly peer review will only pick up the most blatant fraud as it relies on the presumption that the researcher is acting in good faith

The whole thing can land up being a bit shit though as like so many things status, money and pride corrupt. My friend in his PhD tried to build on the research of a well respected academic but couldn't get the results he was expecting. He engaged the academic on his original work and they were initially helpful but then just went silent. My friend was forced to reproduce the original experiment to figure out where he was going wrong and surprise surprise, couldn't reproduce the results. Contacted the academic and eventually the original journal, both refused to engage, presumably as they would land up with egg on thier faces.

I’m entering into CSB territory here but this happened when I did a PhD in the early 90s - not a well respected academic but building on the research of a former PhD student even if the key point was only in passing rather than the basis for their thesis.

Try as I might (and factoring in that I was a bit shit), I simply couldn’t replicate the results let alone build on them but this was apparently all down to user error. Which, given the bit shit bit, I kind of accepted.

Anyway, research went nowhere and it was only later that I found out that the key finding that I had been trying to build on was “not readily repeatable”.

Luckily I saw the funny side of things and, arguably as I was sponsored by BP, it was all karma.

EDIT - this was test tubes, pipettes, lab coats and shit

Yeah that's a big problem, even for some of the most famous studies.
https://www.npr.org/sections/health-shots/2018/08/27/642218377/in-psychology-and-other-social-sciences-many-studies-fail-the-reproducibility-te

WAPO a month ago

https://www.washingtonpost.com/technology/2022/06/11/google-ai-lamda-blake-lemoine/

Yeah he got sacked this week.

https://www.bbc.co.uk/news/technology-62275326

I wonder if we'll (not me, obvs) get AI and self replicating machines up and running before the climate catastrophe wipes us out. #legacy

Elsewhere:

1 Attachment

I missed that.

Some interesting philosophical arguments in the Twitter thread

https://twitter.com/cajundiscordian/status/1535627498628734976?s=21&t=f9n-Af6o43ZaHyJO3cb2cw

Unfortunately as part of the scientific method, peer review often relies on a person to look at a fairly dense piece of information in minuscule detail in their spare time, for free with normally short turnarounds. It’s no surprise if things like image manipulation aren’t picked upon unless they’re jumping out the page

Was said PhD student there to teach you at the start? Most scientific methods are written as impenetrably as possible I’ve found.

For instance “ do measuring procedure to this thing” also involves a 4 hr process involving said thing with an improbable homemade piece of measuring apparatus which is never mentioned in paper.

And then people wonder why science isn’t replicable

.

Don't get me wrong, I am extremely intrigued by AI on the whole. In a few weeks I will be starting a new job, and the fact that this company makes (among other things) an AI tool is the prime reason I am switching jobs. But to me it just sucks that a program that does nothing else than just talk a lot gets all the attention.

12,000 acres burning near Yosemite:

https://twitter.com/realflanbinflan/status/1551229855840325634

This is truly awful. Multiple impacts on the environment and the communities.
There is a feeling (in my mind) that this is the apocalypse. Plague, war, devastation, crop failures.
I last had a similar feeling in the late 80’s when we had several mass transport disasters in a short period.

Presumably this will happen more often due to global warming.

I was in Yosemite after the last lot of fires. I can’t even describe the devastation.

The previous student was not around, no

https://www.theguardian.com/environment/2022/jul/25/tyre-dust-the-stealth-pollutant-becoming-a-huge-threat-to-ocean-life

London narrowly avoided a blackout last week

https://www.bloomberg.com/opinion/articles/2022-07-25/london-s-record-9-724-54-per-megawatt-hour-to-avoid-a-blackout

Expecting 2 days of 37C highs with not much wind. I pray (not literally) that the electric grid is resilient enough to keep up.

Thanks for sharing. A bit disappointed my concerns from 2 weeks ago were warranted. Hopefully there will be a consumption reduction campaign for next year’s (this year’s?) heat wave, especially given the increase in air con units after last week.

Hopefully there will be a consumption reduction campaign for next year’s (this year’s?)

Highly unlikely, at least from those in power, i.e. the mayor / UK government.

Really interesting. And slightly worrying

Last week’s news but interesting- https://www.sec.gov/rules/sro/occ-an/2022/34-95327.pdf

OCC filed advance notice with the SEC intending to increase its non-bank liquidity facility from 1 to 3 billion. It also seeks to remove OCC’s obligation to file advance notice to change its level of required aggregate commitments, instead allowing its Board of Directors to set via resolutions the level of aggregate commitments at a level no lower than its current $3 billion size.

So the people in charge of being prepared to bail out a failed US bank are saying they see the need to have access to more resource given the current situation. The good view is that they’re making sure they’re keeping updated and prepared for any eventuality; a bad view would be that they see heightened risk of a bank failing. Hopefully the former! 🤞

They do say that they have orgs wanting to invest which currently can't because they're at their $1bn limit and they've increased the limits on other funding options so it could just be then trying to maintain balance across all their cash sources.
Then again, with the markets doing what they've done and inflation kicking about I'm sure a fair few financial wizards will be uncovered as snake oil sellers.

The heat also caused a few cloud servers to shut down.

https://gizmodo.com/google-oracle-cloud-heatwave-europe-1849199124