nealmcb2

You can run Galactica (the "base" model) for free on Colab (https://colab.research.google.com/). It takes about 4 minutes to start up. Just specify a "GPU" runtime on colab, and follow the simple instructions from their github (https://github.com/paperswithcode/galai):

!pip install galai

import galai as gal

model = gal.load_model("base")

model.generate("Scaled dot product attention:\n\n\\[")

I love Kim Zetter's tweet:

Must have been a turnip truck https://twitter.com/KimZetter/status/1384936715769503745

That's a dig at the turnipesque savvy of Cellebrite, which royally stumbled into this.

cf. https://www.dwt.com/blogs/privacy--security-law-blog/2015/03... > FCC chair Wheeler said the FCC “didn’t just fall off the turnip truck.” Through CALEA, CPNI, and CSRIC, Wheeler said the agency has been working to protect consumer privacy.

The best overview of this mess is from Cory Doctorow, who notes that while the claims of fraud in this election are garbage, the voting machines, and especially overused Ballot-marking devices (BMDs), are also hugely problematic.

Voting machines didn’t steal the election. But they’re a terrible technology. - The Washington Post https://www.washingtonpost.com/outlook/2021/02/03/voting-mac...

And worst of all is ES&S' threats against folks who point out those legitimate flaws, e.g. SMART Elections, a journalism and advocacy group in NY.

ESS voting machine company sends threats https://freedom-to-tinker.com/2021/01/11/ess-voting-machine-...

What we need is evidence-based elections:

Evidence-Based Elections: Create a Meaningful Paper Trail, then Audit Andrew W. Appel* & Philip B. Stark* Georgetown Law Technology Review 2020 https://georgetownlawtechreview.org/wp-content/uploads/2020/...

And though we have improving evidence for correct tabulations in the recent elections, many states still aren't doing a good job providing evidence for ballot eligibility or chain-of-custody.

The best way to promote confidence in our democracy is to calmly stick to facts, not be swayed by politics, and work to make things better. For more on how we've done that in Colorado and elsewhere, see my page at http://bcn.boulder.co.us/~neal/elections/corla/

That article makes good points, but doesn't even mention the obvious alternative: join the fork that Amazon immediately announced, the need for which has also been supported by other significant contributors besides CrateDB like Logz.io and Aiven. See https://www.zdnet.com/article/aws-as-predicted-is-forking-el...

Re: > it was to inhibit Cathedral-style codebases from using GPLed software at all.

I'm surprised at the notion that RMS ever had the intention of inhibiting GPL use in Cathedral-style codebases. And I agree with others that big frameworks existed back then also.

See also:

https://www.reddit.com/r/reinforcementlearning/comments/iyk2...

An adaptive deep reinforcement learning framework enables curling robots with human-like performance in real-world conditions

Apparently that isn't always the case. See the PyPy Status Blog: PyPy's new JSON parser https://morepypy.blogspot.com/2019/10/pypys-new-json-parser.... which talks about being more efficient with both deserialization and memory.

Compliance with SEI CMMI Level 5 probably is expensive. While looking for more data I noted that several Boeing units are at Level 5. And I also found that between 2008 and 2019, about 12% of appraisals given across all industries and countries were at maturity levels 4 and 5, and hundreds are at Level 5 (more than Level 4), according to CMMI Institute: https://cmmiinstitute.com/resource-files/public/cmmi-adoptio...

So it all makes me wonder about the official maturity of Boeing's 737 team, and also how much that rating relates to actual software quality and safety objectives (as apparently achieved in the space shuttle work)....

The raw ballots are available at https://www.debian.org/vote/2019/vote_002_tally.txt

It would be helpful to also have that shared in in a way that makes it easier to see how many top-ranks each choice had, how many bottom-ranks etc. But note that it seems that equal rankings are allowed, and many common ranked-choice cast vote record formats don't represent equal rankings.

That is an excellent and very helpful writeup!

I'm particularly disappointed to hear that Google doesn't provide any way to rotate the signing key for an app. Is there an issue for that file with them anywhere, or more discussion?

Some day, I hope reputable services have migrated to The Update Framework, which has been pointing out and solving these and other problems related to software updates for several years now.

https://theupdateframework.github.io/

Actually, a quick search leads to this - is it indeed possible to rotate your key, at least for Android's Pie version?

  https://www.androidpolice.com/2018/08/13/android-pie-includes-key-rotation-way-developers-change-app-signatures/

The point is that they set the toll price to eliminate the bottleneck, i.e. to get the traffic flowing at 45 MPH or more. The previous bottleneck was caused by solo drivers who just ignored the HOV2+ restriction. So you have a new choice, to pay and get a much more efficient and predictable commute.

That expresslanes URL is for I-495. The I-66 historical data is here: https://vai66tolls.com/ And I too would love to see that data in a more convenient form, via an API, etc.