axg11

22 ms·

181.
by axg11 4y ago
> I have serious trouble seeing how “collective agreements” has any intrinsic value. Then you probably don’t see value in any form of cryptocurrency. I do, because I think collective agreements are at the heart of how society functions.
182.
by axg11 4y ago
Calling it a Ponzi scheme is uninteresting and not really the point here. Was GameStop a Ponzi scheme when the stock price soared? I think the interesting part here is the relation to bitcoin and a new theory on how to price things. It’s tr
183.
by axg11 4y ago
What does nanotech mean to you?
184.
by axg11 4y ago
This is a great point - too many companies forget to optimize for speed of UX. My theory is because speed of UX isn't a strong selling point to decision makers. JIRA sells well because it's a safe choice (recent fuckups not-withst
185.
by axg11 4y ago
The most important point in this article is that whether Tether is fully collateralized ultimately doesn't matter. As long as there are well-capitalized parties (Bitfinex, other exchanges) that want to prop up Tether, it will be fine.
186.
by axg11 4y ago
See also RETRO, a type of retrieval transformer: [0], [1], [2] [0] - https://www.deepmind.com/publications/improving-language-mod... [1] - https://jalammar.github.io/illustrated-retrieval-transformer...
187.
by axg11 4y ago
The best software is developed by people who have deep knowledge about the use case and/or the ultimate customer. What proportion of PMs at health software companies have worked in a hospital? I'd wager it's vanishingly small
188.
by axg11 4y ago
Compare the UK vs. US [0]: life expectancy is higher in the UK _and_ expenditure is lower. As much as people love to dunk on the NHS, population-level optimization works. [0] - https://ourworldindata.org/grapher/life-ex
189.
by axg11 4y ago
This is the reality. In fact, the situation is getting even more complicated as direct-to-consumer screening becomes more popular (e.g. full body MRI scans). Companies that are selling the health screenings are only incentivized to sell as
190.
by axg11 4y ago
This is a positive development and validates mRNA vaccines as a general platform. Having said that, I'm pessimistic about the impact of vaccines on COVID. Despite a _gigantic_ global coordinated effort, most people on Earth are still u
191.
by axg11 4y ago
This is great. I love how this allows you to quickly compare and contrast different cities. London and Vancouver are very similar, and both very different to Orlando. My only critique is that without showing precipitation these visualizatio
192.
by axg11 4y ago
What is the SOTA for computational models that predict enzymatic activity?
193.
by axg11 4y ago
What makes you think any chip would be harmful to ingest? You could happily eat most chips, if they were small enough, without any negative effects.
194.
by axg11 4y ago
What does pandemic preparedness mean in this context? Most countries that fared badly due to COVID did so because of policy decisions, not lack of materials. We could invest in faster vaccine development, but again the uptake of vaccines in
195.
by axg11 4y ago
This move is essential for Plaid to survive. Stripe just recently moved into Plaid's core business and Stripe are increasingly becoming a one-stop shop for all company needs that involve money. If you already have a Stripe account for
196.
by axg11 4y ago
Web3 is a loosely defined trend and every person has different expectations for the changes that web3 should bring. Here's what I personally think web3 can do to change the internet for the better: - Increase proportion of consumer spe
197.
by axg11 4y ago
There are a few SaaS templates out there and services such as Memberstack that make it easier to implement the boilerplate SaaS features in Webflow. I'd also recommend just doing the boring work once and keeping that as a template for
198.
by axg11 4y ago
USD collateralized stable-coins are happening whether central banks participate or not. I think the advantage for a central bank would be to bring stability and increase demand for the underlying asset (the centrally controlled currency).
199.
by axg11 4y ago
I agree that 1% (if true) is not insignificant, but destroying all cryptocurrencies to reduce our energy consumption by 1% isn't a clear tradeoff to me. We also need to account for the fact that renewable energy as a percentage of ener
200.
by axg11 4y ago
Aside from optics, why doesn't a major central bank step in and create a stable-coin? A GBP/CAD/USD stable-coin collateralized by a central bank's reserves would be widely adopted at this point.
201.
by axg11 4y ago
A great, laconic statement to rally the anti-crypto troops. In reality, destroying cryptocurrencies would have zero effect on any of those three: Estimated total energy usage for all cryptocurrencies: 0.1-1% of global energy usage Inflation
202.
by axg11 4y ago
Race _is_ a spectrum but genetic differences themselves are distinct (SNPs). It's trivial to train a classifier to distinguish race from genetic data, hence, I'd argue they are distinct groups. You can draw an analogy to colours i
203.
by axg11 4y ago
You are describing most successful machine learning models. Take AlphaFold, it has surely discovered relationships that govern protein folding better than any human has ever previously understood.
204.
by axg11 4y ago
Perfect example of citations-driven research. The authors aren’t motivated by a genuinely interesting scientific question (“are anatomical differences between genetically distinct groups of people visible in X-rays?”). Instead, the authors
205.
by axg11 4y ago
I used to work in HFT. I promise you that companies would still try to exploit randomized batches. There is an advantage to being the very last entrant into a batch (most up to date information). Truly random batches are not trivial to impl
206.
by axg11 4y ago
> ML models are great tools, but they're way too much of a black box. A human doctor is also a black box, in meat form.
207.
by axg11 4y ago
That sounds like a complex solution. Sometimes a dumb solution that works good enough is better than a complex solution that _probably_ can't be exploited.
208.
by axg11 4y ago
A question for the computer scientists here: are there applications for this where Codex could be used to automatically optimize complex systems?
209.
by axg11 4y ago
100% agreed. I already see myself doing this with Github Copilot. If I write a comment or start a line of code in a certain way, I get a much better suggested code completion.
210.
by axg11 4y ago
> "That was my first near-accident on beta" As a tech and machine learning maximalist I'm frustrated with how haphazardly Tesla is approaching self driving. This is not a technology that should be QA'd consumer-first.

More ›