potamic

How did he not share code if you're working together?

> our policy agent extracts all coverage limits and policy details into a data ontology.

Are they using some software for this or was this built in-house?

Humans are still needed, but they just got down-skilled.

Got a source?

> It is a luxury good that only a very small number of people care about

The world runs on computers. It is as essential as oil for the functioning of societies. Increase in silicon costs is going to increase costs unilaterally across the board. It happened during the pandemic and something similar will happen now. If anything it should be a wake up call to countries to start thinking about securing their own supply chains.

Would it be as ironic if one wrote an article about Google and used Google search to research their sources?

One could say the same thing about a higher power as well. But the scientific stance is clear on that one.

That's not how science works. You have to devise an experiment that can independently validate the assertion.

It's not good enough for science. How do you approach it so it can be independently verified?

But is consciousness even a thing? Shouldn't the burden of proof be on the one making a claim that there exists something called consciousness? If they cannot show evidence that such a thing exists or may exist, then for all purposes it does not exist.

When all you have is a hammer, everything looks like a nail.

Are you sure 2 minutes? That's 1000+ tokens/sec.

What's the token usage and time taken to run this for a given repository?

6 year time period between 1990-1996. Would have been interesting to see how these results would have fared across the 2000 crash.

But I don't think the paper claims that they beat the market. What they are concluding is that you can classify investors as skilled and unskilled, with the skilled group able to achieve returns consistently. But whether those returns can beat the market, they say the evidence is limited.

Well, the last 5 years or so is the epitome of a bull market.

Did the paper say the same set of people consistently beat the market over long periods of time or did it say at any time about 10% of investors happen to beat the market?

For someone who developed their own quantum processor they're pretty light on technical details. Anyone knowledgeable in this field chime in, if this is indeed the case or is it likely white-labeled from another manufacturer?

Not a dentist, but my read of the situation is that dentists generally are not very excited about doing fillings and there's a push towards getting into more complex procedures like root canals, invisalign and implants. It's probably partly due to wanting to upskill and increase your repertoire and partly due to the margins. The margins with these procedures can be an order of magnitude more than that of fillings, especially anything that is supplied by a big brand.

Isn't ART also destructive? Just that it's done with hand instruments instead of a drill

Do you have any examples/demos for very large codebases? Because it's easy to get decent quality output using LLMs on small codebases. But large, messy codebases is where you really need help with understanding.

Why would someone do that? Especially for presidency which is the final stage of their career? They're not beholden to or reliant on anyone no more so shouldn't have to be swayed by any adverse interests.

How did you test this? Did you train something?

Is there a reason to believe that consciousness, sentience and emotions exist?

That's kinda scammy. It's not like they have to manage shipments and handle goods or anything. I wonder if they're banking on a percentage of users leaving credits unused like credit card companies do with loyalty points.

What if one doesn't own an android/iphone device? Banking is a fundamental need, so most countries regulate them to cater to a wide range of users. In this case it's possible that the bank could be compelled to provide you a 2FA device if you don't have one.

The most important principle in software architecture is that no principles are absolute, except this one, otherwise it would be a paradox. Theoretical correctness must be balanced by practical necessity. I don't know what the ulterior motives were in this case, but I can think of many possible reasons. Unreliable, lack of support, lack of tooling, missing observability, operations heavy, deprecation plans, dependency on another team that is unable to cater, etc. there could be many reasons one might prefer to trade-off a small performance penalty against these. Trade-offs are unavoidable in any significant system, and this seems is one of the smaller ones you would see IMO. What's more important is to make sure the trade off is a conscious choice and to plan for the consequences.

It's sad that such a thing needs regulation in the first place. In real life if a salesman is being inconsiderate, I'll go out of my way to avoid their sales and find someone else who is better mannered. But we don't seem to apply the same measure to ads. Ads can be brash, insulting and manipulative, and yet that doesn't seem to cause a negative outcome for them. Rather it appears such ads work better and now that's what everyone's pushing towards. Human psychology is such a weird thing.

It does mention compliance with the CALM act, which lays out the precise methodology by which loudness will be measured [1]

> The Calm Act refers to A/85, and A/85:2013 specifies BS.1770 (specifically referencing BS.1770-1) as the source of its loudness measurement techniques (1770-2 did not exist at the time A/85 was finalized). So BS.1770-1 currently serves as the yardstick by which U.S. television programming will be evaluated for CALM Act compliance.

> BS.1770 recommends the Leq(RLB) measurement algorithm, where Leq(W) the frequency weighted sound level measure, xw is the signal at the output of the weighting filter, xRef is the reference level, and T is the length of the audio sequence.

> The drawback of BS.1770 as originally conceived is that it measures average loudness over the entire length of content. This may be fine if the loudness is fairly consistent over time. If not, a quiet section of content may, as illustrated in Figure 5, bias the average level so that it measures as acceptable despite having some sections that are unacceptably loud.

[1] https://www.telestream.net/pdfs/whitepapers/wp-calm-act-comp...

Why isn't there something similar for running OSX apps on linux? Should be easier than wine in theory because it's unix based?

What are those 1000 repos? Not sure I understood the relevance to this post.

Sure, any significant business depends on a lot of software. As a software company you don't want to build that which is not in your core business, so you would buy a lot of other software. But the reality is a little more insidious. Investors are mostly selling to themselves. Every time an investor invests in a company there is a push to partner up or avail services from other companies their portfolio. Even if it's not the best choice for the company at that time, investors are more interested in strengthening their overall portfolio.

The long term consequence of this is that as capital gets concentrated into fewer and fewer hands, so does the means of production. If you are outside the network you will find it very hard to sell your services even if you have a better product. This already happens in many places with wealth disparity. Deals simply do not happen without the "blessing" of a wealthy backer. Software is unfortunately moving that way. The once upstart nerd who was able to launch their business from the basement has turned into the technocrat baron who will keep such a thing from happening again.