It's almost like he's undermining the one single thing the diamond sales industry had used to thrive for nearly a century before him.
5,689 karma · joined December 29, 2013
But seriously, Python and JavaScript have reflection/introspection features. Once the investment was made to transcribe the entire language into a compiled language, the speed gains were obviously going to happen.
Arguably, though, the US has been "stealing" (in the form of sanctions and frozen assets) from other countries going back more than a few decades, at least as far as Iran in 1979, so your timeline doesn't hold up.
"Stealing" is wordplay when it comes to Russia because they are at war and the things that are "stolen" are assets being frozen to help pay restitution for the war they started. And Russia would deny the oil stolen from their ships was theirs because the entire point of the "shadow fleet" is to obfuscate ownership, mix and match the fluid products, to create plausible deniability to avoid sanctions.
There is a huge difference between a guaranteed Constitutional right that this country fought a Civil War for... and a privilege such as getting on a plane or driving a car.
Assume that I know everything you know about this topic and still arrived at a different conclusion.
There's no reliable way to verify a foundation model has been trained on a particular piece of proprietary data. If an API key is ingested, hopefully the foundation model is wrapped in enough moderation that the raw API key oberserved during training is not recited verbatim in the output.
The point of the government doing it is that it has to be able to service those people who live maximally far from a DMV (say 60 minutes each way by car), who can't drive there, who have to weigh losing pay to take the time to travel there, who have below average intelligence, who are functionally illiterate (because Mississippi created A LOT of those before their recent education reforms), who "can't use technology", who forgot to bring whatever specific document qualifies for Real ID, etc.
Like I said, the USPS is in a much better position to be able to scale identity to the national scale.
Imagine all of the blast radius to society if this type of incentive is reproduced everywhere. Police searches through Flock databases, matches for hiring candidate resumes, organizing targets in a war with Iran, etc.
It's like someone reading a National Enquirer article about "Hillary Clinton being an alien from outer space" (a real headline topic from decades ago) and drawling the conclusion that all journalism is "a lie". The user has to understand media literacy and be at least a little skeptical of the claims that are made, then cross reference with another source.
It wasn't "catastrophic" for the largest of the 3 US credit reporting agencies when their entire dataset was breached. The company is 100% IP and the only value they have was completely copied. Their largest value is to verify identities by the things Americans know (KDB) and after that "single factor of identity" was 100% compromised, the company only got bigger and more contracts.
When there are only 4 competitors in the large scale foundation model business and they all throw caution to the wind because they are racing to own the "$30 trillion TAM" they are all going to make critical security, RBAC, and segregation mistakes.
Both ChatGPT and Claude threads marked for sharing have been indexed in Google at large scale. This is incredibly easy to tell Google crawlers via robots.txt not to crawl those URLs, but nobody at either of these uber unicorns could be bothered to add that one pattern to the one file.
And all of the skepticism here is about verifiability. The foundation model companies are liable for potentially more the companies are worth if found to be violating copyrights of content used for training. They aren't going to make it easier for lawsuits against them by detailing their data ingestion into training pipeline.
She's just arguing that most of the purposes of code reviews should be done before the typing part of coding instead of after. Identifying the best design before investing in typing / tokens.
Waterfall has to do with the size of the loop and when the customer gets to provide consumable feedback. She's not proposing changes to the size of that loop.
She's arguing that most code reviews aren't necessary. Some can be automated if they are deterministically predictable (eg. formatting, lint, standardization changes). Others should be reviews of design before the coding phase.
She gives a few specific examples of when code reviews should be maintained.
S Korea has had government issued digital IDs for all online transactions, although it was originally implemented as an Internet Explorer 5.5 plug-in (eww). I haven't heard anything about how it works these days.
Japan has new digital IDs cards and a standard little device available at every corner store to use it to digitally sign legal documents.
It's worth pointing out that the point of contact for state IDs is the DMV, which is the butt of every government inefficiency complaint. If, instead, it was done at USPS offices or even by postal workers on their routes, no one would complain.
It's entirely about whether the burden is placed on the citizen to maintain their Constitutionally guaranteed rights or whether political agents can use the state to selectively burden neighborhoods, especially ones with no / badly performing DMV offices.
The CEO (or at least the founder) created and cultivated the culture of the company thus far. If the culture is bad, they are the single largest reason why.
You are right in that execs won't seek out the blog article which criticizes their strategy... unless they actually want to steelman against their own position (which is a very useful activity but few people do it).
But the CEO doesn't need to find it on their own. They surround their self with a Board, investors, sales people, managers / leadership team, people from their MBA program, etc. the article just has to plant a seed which grows somewhere in that ecosystem.
That said, the article may not be right. The largest and most successful companies in the world aren't the ones with the best company culture. They simply work in industries where the money is so prolific that the people who want to work in the industry are already highly motivated. The CEO doesn't need access to this information because it isn't true; it simply tickles your cognitive biases.
Ignorance. I'm not a Luddite. I use LLMs and see their relative value.
I'm trying to get idiots who think "all gas, no breaks" is a reasonable strategy to recognize there are consequences to ignoring the voices/will of a massive chunk of society.
The fuzzer found the bug before any humans did, so there is a mismatch of developers who could find this bug and those who did (without an LLM-coded fuzzer).
The value of the fuzzer continues long after it found this one bug.
It's worth nothing that in the bug discussion thread, the bug fix author pointed out that it's not easy to set up the config then call the functions in the right order. Your comment assumes that the reader has enough context to read the code and build the finite state automata in their mind. The bug fix reporter's comments suggest that you are assuming things which you shouldn't assume.
If that were truly a problem, why wouldn't a human-only social media platform take off like wildfire?
Do you honestly believe that the ledger for automation has zero items in the "liabilities" column? Do you think that 100% of people whose job is displaced find another job in a reasonable amount of time and on the same order or pay / status?
Do you believe that neighbors of recently built data centers have nothing to complain about?
The truth is that "data center politics" is happening in the context of post-NAFTA offshoring of lots of jobs/industry, post-manufacturing migration to the South, post-hollowing of labor unions, and in the context of Trump's War on Immigration (where AI companies have likened their product to "infinite AI immigrants").
Even the White House's Ratepayer Protection Pledge (a laudable goal) event was undermined by POTUS's speech at the event. Instead of even pretending to listen to the people who are hesitant about first and second order effects of data center buildouts, he encouraged developers to push the projects through ASAP and ignore all of the naysayers. It's clear this entire project is to maximize commercial advancement, not benefit people or give voice to voters/constituents.
And in a moment where we are told we have to do this ASAP because "we are in a race with China" but nobody can explain the race. It's not for AGI because China doesn't expect it to yield AGI. Perhaps Zuckerberg does, but he was wrong about crypto, then Metaverse, then open weights models training (before he returned to open weights). The US seems to be using LLMs wrong as a glorified DEI detector (DOGE using LLMs to enforce their political homodoxy) and to identify schools in Iran to bomb + double tap.
China has reasonable regulations on LLMs to protect their society whereas our "regulations" seem to be the Amazon CEO texting David Sachs on a Friday evening to warn about cybersecurity in a misunderstanding between what Mythos (prerelease for select privileged users) and Fable (general release with lots of moderation controls) and the nature of the cybersecurity vulnerabilities found (they were able to be replicated with other non-Claude models).
We are being outwitted by ourselves. The problem isn't the people anxious about data centers or AI. It's the clueless leaders who don't bother to anticipate the obvious outcomes of their decisions and execute a plan to mitigate it. And you desire "feeling awe" so much that you don't want to recognize that politics means not everybody is going to succumb to the same feeling of awe, and that's okay.
Investors know the US is going to lose this status if the current momentum continues and there are some possible alternatives.
Investors won't sit by and tolerate MONEY PRINTER GO BRRRRR if they have ANY alternatives and they have LOTS of incentives to find any.
https://x.com/macropaperr/status/2089255200918007854?s=46
Obviously we don't measure it in "how many traders", but what the effect their trading has on the instruments they buy.
The US Dollar is slipping from the world's reserve currency status and the US treasury is slipping from the default safe haven for liquid assets. The BRICS countries bonds are becoming more attractive as of this month.