HNHacker News
TopNewBestAskShowJobs

riemannzeta

447 karma · joined May 29, 2008

submissionscomments
riemannzeta··on The pivot table, the spreadsheet's most powerful tool (2020)
Something similar happened to Stac Electronics. https://en.wikipedia.org/wiki/Stac_Electronics

But Stac had patents, lawyered up, and sued Microsoft for patent infringement. Microsoft didn't lie down, but after a court ruled that both of them needed a license to ship products, Stac went to the largest OEM customers and offered a license. It was either take the patent license or not ship, and the pressure from the OEMs on Microsoft resulted in a settlement that was not terrible for Stac.

This is one of the best examples I know of how software patents can be good for competition.

riemannzeta··on Nvidia’s AI supremacy is only temporary
That's fair. It seems we all agree that the timeline here is much longer than some might understand from the author's remarks. But ultimately I agree with the author regarding Nvidia's competition — it's like a dog walking on its hind legs: it's not done well; the surprise is that it's done at all.
riemannzeta··on Nvidia’s AI supremacy is only temporary
I agree, if by "long" you mean 5 years. The author seems to have a longer time horizon in mind than do most Wall Street analysts though.

Sun Microsystems did great through the 1990s. But by the time the LAMP stack was fully mature, web servers had been commoditized.

riemannzeta··on Kullback–Leibler divergence
I will confess that at least the biology work is still somewhat mysterious to me. John Baez used to be active on Twitter. I believe he's still active on Mathstodon now. If you're curious, you might ask him.
riemannzeta··on Kullback–Leibler divergence
KL divergence has also been used to generalize the second law of thermodynamics for systems far from equilibrium:

https://arxiv.org/abs/1508.02421

And to explain the relationship between the rate of evolution and evolutionary fitness:

https://math.ucr.edu/home/baez/bio_asu/bio_asu_web.pdf

The connection between all of these manifestations of KL divergence is that a system far from equilibrium contains more information (in the Shannon sense) than a system in equilibirum. That "excess information" is what drives fitness within some environment.

riemannzeta··on HashiCorp adopts Business Source License
Somebody should point out to them that there's an error in their Parameters:

https://www.hashicorp.com/bsl

The "Licensed Work" parameter should refer to what they are licensing. Right now it reads "The Licensed Work is (c) 2023 HashiCorp, Inc."

I don't see how a corporation itself is copyrightable content to which a license may be granted.

riemannzeta··on Zluda: Run CUDA code on Intel GPUs, unmodified
Good luck enforcing those software parents though in the current environment. My sense is that hardware companies put more faith in the enforceability of patent claims to something like CUDA than software companies with experience with litigating patent claims to APIs would. Post Oracle v. Google, something like CUDA is vulnerable to being knocked off.

And FWIW, that seems to be a reasonable result given the overall market structure at the moment. Having all eggs in the Nvidia basket is great for Nvidia shareholders, but not for customers and probably not even for the health of the surrounding industry.

riemannzeta··on ChatGPT app for iOS
Sigh. Another iOS only launch.
riemannzeta··on Five Books: The best books on everything
Would be more interesting/useful to see a meta-list ranking the books that come up more than once in recommendations, sorted from most frequent to least frequent.
riemannzeta··on I'm never investing in Google's smart home ecosystem again
The ultimate question for consumers though is this: Are you better off in the Apple ecosystem? As bad as things are with Google, the answer is probably no. But not many have the time, tools, and training to be independent of either.
riemannzeta··on I'm never investing in Google's smart home ecosystem again
Not allowing good search within a channel is the most ridiculous. They're a search company. How is that self-consistent? Especially when you know they have the data and index, but are simply using it for advertising.
riemannzeta··on I'm never investing in Google's smart home ecosystem again
Their position in the Sonos lawsuits is looking pretty stubborn and selfish at this point. Just pay for a license rather than inconveniencing your customers. They should be working together rather than fighting over a royalty that's relatively small to Google and huge to Sonos. Seems stupid to me.
riemannzeta··on Prediction and Entropy of Printed English (1950) [pdf]
According to Thorp himself they were collaborators on the roulette scheme.
riemannzeta··on Prediction and Entropy of Printed English (1950) [pdf]
The full story of the invention of the first wearable computer is told in Ed Thorp's memoir https://www.amazon.com/gp/aw/d/B00SEFEYCI/

Highly recommended!

riemannzeta··on A trick to reaching flow: Leave your work broken
“Stop in the middle of a sentence, leaving a rough edge for you to start from the nest day – that way, you can write three or five words without being “creative” and before you know it, you’re writing.” — Cory Doctorow
riemannzeta··on Clocks and Causality – Ordering Events in Distributed Systems (2022)
Yup. Different problem space! Thanks for laying it out clearly.
riemannzeta··on Clocks and Causality – Ordering Events in Distributed Systems (2022)
Yes, 1 seems to be the most important difference, with 4 as one of its most important consequences. It's definitely intended to be a solution to Byzantine faults (or double payments as they would show up in practice if the transactions are financial!).

I think your summary in 2 is accurate. The answer to scalability 3 seems to be that verifiable delay functions are always costly to solve the first time, but can be checked in parallel -- like a maze that takes time to solve the first time, but can be checked instantly simply by looking to see whether the solution crossed cross any walls.

It seems to me like a reasonable technological solution to time-ordering transactions in a decentralized ledger subject to byzantine faults, but I'm not an expert here. It's at the core of the Solana layer 1 blockchain. Solana took it on the chin with the FTX collapse because FTX had chosen Solana as its main exchange and made a heavy investment in its ecosystem. Wondering whether that was an exogenous shock that makes Solana a relative good investment opportunity at the moment. The capability of doing verifications in parallel means verification of 100k transactions per second. Which is on par with what the credit card companies can do.

riemannzeta··on Clocks and Causality – Ordering Events in Distributed Systems (2022)
Curious to know whether, in your opinion, the use of verifiable delay functions and proof of history represents a more reliable and scalable approach to decentralized time-ordering.

https://solana.com/news/proof-of-history---a-clock-for-block...

riemannzeta··on Synchronization and Banking
Synchronized deposit withdrawals are a positive feedback loop that results in a market failure. More generally, market price signals help to stabilize cycles of consumption and production.
riemannzeta··on Google releases Bard to a limited number of users in the US and UK
This is a test for theory of mind -- a term of art used by psychologists for the ability to understand that other people have their own consciousness, with different memories and awareness.
riemannzeta··on GPT-4
Is anybody compiling a list of errors specific to GPT-4?

This has been a great resource to-date:

https://github.com/giuven95/chatgpt-failures

riemannzeta··on Launch HN: Pynecone (YC W23) – Web Apps in Pure Python
Your ideas seem really cool. I'm sad that you chose a name so close to another great startup that does vector databases and search. https://www.pinecone.io/

You're headed into a place where some customers may confuse you.

riemannzeta··on Quitting the rat race
I believe Mr. Barry is simply doing what very few of us are brave enough to do: To think for himself!

We are social animals, and many of our goals are the defaults set for us by our social cohort. Very seldom do we take the time to think deeply about what we really enjoy, much less have the courage to act on what we might think in a way that would mark us as different. After all, any rejection of the defaults is a subversive threat to undermine the entire system!

Good for you, Mr. Barry. I for one am very excited to hear more from you in the future!

riemannzeta··on US Department of Energy: Fusion Ignition Achieved
Great comment. But you meant 2 helium atoms and 1 neutron in the second equation, correct?
riemannzeta··on Geniuses of the past were aristocratically tutored
Was he tutored? I also thought of Ed Thorp. But he was NOT tutored. He was kind of an autodidact according to his memoir. He had good teachers too, but not aristocratic tutoring by any stretch.
riemannzeta··on Geniuses of the past were aristocratically tutored
I believe at least part of the answer is that many of the would be geniuses went into financial services. Ed Thorp and Jim Simons come to mind as examples of people who made important contributions in mathematics, but spent a substantial chunk of their lives figuring out how to take money from people who are dumber than they are.
riemannzeta··on US PTO rules against Univ of California, argues that Broad Inst invented CRISPR
Actually, one of the fascinating facts about the history of PCR is that Kary Mullis didn't optimize to make billions on the patent rights. In his Nobel acceptance speech, he reminisces about how a coworker told him to quit his job and start another company before submitting an invention disclosure on PCR, but he didn't do that.

https://www.nobelprize.org/prizes/chemistry/1993/mullis/lect...

My friend Ron Cook, who had founded Biosearch, and produced the first successful commercial DNA synthesis machine, was the only person I remember during that summer who shared my enthusiasm for the reaction. He knew it would be good for the oligonucleotide business. Maybe that’s why he believed it. Or maybe he’s a rational chemist with an intact brain. He’s one of my best friends now, so I have to disqualify myself from claiming any really objective judgement regarding him. Perhaps I should have followed his advice, but then things would have worked out differently and I probably wouldn’t be here on the beach in La Jolla writing this, which I enjoy. Maybe I would be rich in Tahiti. He suggested one night at his house that since no one at Cetus had taken it seriously, I should resign my job, wait a little while, make it work, write a patent, and get rich.

riemannzeta··on US PTO rules against Univ of California, argues that Broad Inst invented CRISPR
UC was first to file some other claims. In the claims that were at issue in this decision, Harvard was first to file. Actually the opinion does a good job explaining some of that. But Harvard is labeled the senior party in the opinion because they were the first to file this set of claims.
riemannzeta··on US PTO rules against Univ of California, argues that Broad Inst invented CRISPR
Yes, just in the US, and only one of the sets of patents in the US. This one is about use of CRISPR-Cas9 in a eukaryotic environment specifically. Both parties have some other claims, and ultimately will have to cross-license each other for the technology to develop. Too bad such an important decision is being left to such a small group of university administrators who know very little about how best to grow the technology into successful therapies and technologies at scale!
riemannzeta··on US PTO rules against Univ of California, argues that Broad Inst invented CRISPR
I agree that publicly funded research should be easier to get access to. Ironically, you have it backwards. Before the Bayh-Dole Act, there tended not to be any patent filings on publicly funded research at universities because it wasn't clear what claim the federal government had to the patents. Bayh-Dole, by clearly vesting patent rights in the universities, allowed the universities to take title to the patents and license or sell them to private corporations.

The usual story that gets told here is of fantastic success. Look at all the startups that have come out of university research. Look at all the licensing revenue universities have generated from patents that they own thanks to Bayh-Dole.

I'm actually a Bayh-Dole skeptic. Like you, I tend to believe that too much publicly funded research does get concentrated into certain private hands thanks to patent rights on federally funded research. But strengthening the Bayh-Dole Act would only encourage more concentration.

Some startups need exclusive rights to succeed, and pharmaceutical or biotech startups are the prime examples. But most startups don't need exclusive rights to succeed, and I don't see why we should trust university tech transfer offices to make good decisions about who should have access to federally funded technology and on what licensing terms when that is the case, as it is so often. A weaker Bayh-Dole Act would be more helpful than a stronger Bayh-Dole Act in that regard.

But even some biotech inventions like CRISPR-Cas9 or mRNA vaccine tech doesn't seem to me to be stuff that the federal government should allow to be controlled by university tech transfer offices. Some technologies are too important for that. That's why the Bayh-Dole Act includes a provision for "march-in rights." The surprising fact, however, is that they have never been used, even during the pandemic. That to me seems like the easiest and best place to push the federal government if you feel like too much tax payer funded research is being held hostage by tollbooth workers in university tech transfer offices.

← PreviousPage 3 of 6Next →