Where's the special admission program for lottery-winners, con-artists and pickpockets? Those also function in "the real world" - so why not at Stanford?
In the real world, individuals can't do much. It's only through the collective cooperation and the trust behind such cooperation that allows things to happen. Social Elites come with a wealth of trust from the legacies of families and connections that slowly built them up over hundreds of years. And such bonds survive even without the state, predate it and ultimately build it.
That is the "real world". Everything else is just an abstraction, propped up by a system that has only existed for a definite period of time and will not exist outside of that.
I can't help but notice the contrast in the tone (and content) of this HN discussion, compared to the one on the ruling that ended affirmative action[0] for university admissions. Then, the majority of commenters were on the side of meritocracy. HN is consistently pro-elite, perhaps because a good chunk of folk here see themselves as intellectual elites.
> How can you get overlapping clusters if the two sets of labelled examples are disjoint?
What's disjoint are the training labels and the classifier's output - not the values in high-dimension space. For classification tasks, there can be neighboring items in the same cluster but separated by the hyperplane - and therefore placed in different classes despite the proximity.
wut? It was Trump[1] who invited the Taliban to Camp David, negotiated with them sans-Afghan governenment, and started the process of withdrawal with troop reductions, a deadline and everything.
Absolutely, and it was Bush who started it. But Biden oversaw the retreat and it was our biggest failure since Vietnam, and so it will always be his loss. This is also why LBJ is 'LBJ's war' even though he, too, did not start it. Trump's well aware of this reality which is why every interview he does he tries to stress that Ukraine was Biden's war, but he knows that in the end he inherited this disaster and so, in the end, he'll be the one associated with it, so he wants to 'cleanly' wash his hands of it as quickly as possible. And since Zelensky seems increasingly delusional, it's likely that giving it to Europe is his only real out.
Wouldn't you find it strange for a co-lead of an Ethical AI team to not to do any AI safety work? I realize the AI Doomer vs AI Zoomer is a culture-war with a veneer or technical jargon, but I hope we can at least agree on the basic tenets of our shared reality even as we draw wildly different conclusions.
> There will simply be no incentive for companies to develop new video compression technologies, at very significant cost because of the sophistication of the field, knowing that their assets will be thankfully – and nothing more – accepted and used by AOM in their video codecs.
I don't think he fully considered the motivations of Alliance members like Google (YouTube), Meta and Netflix and the lengths they'll go to optimize operational costs of delivering content to improve their bottom line.
> This is what OP meant with multiple coding agents.
It may be the same coding agent behind the GHA. I question the implicit declaration behind OPs critique: that all 160,000+ Google folk should offer a single coding agent to their billions of users (or whatever the TAM is for coding agents). This is akin to criticizing Google Cloud for having VM, Kubernetes clusters and AppEngine; superficially, these products solve the same problem.
FWIW - this Github Actions integration is close to my ideal AI agents workflow[0]. I don't want to metaphorically look over my agents shoulder as it works in a specialized, vendor-locked IDE. I want agents to work asynchronously, taking however long they need, and tackling multiple tasks, with PRs/CLs as the unit of work. Current models may not be up to the task of single-shotting this, but the task is parallelizable across multiple agent-instances & the best solution selected (climate change be damned). I suspect Github alone may not provide adequate context as it may be missing previous tickets and design documents & the back-and-forth on requirements, but it's a start, and I'm glad Google is exploring this path for agents.
0. I believe in this workflow so much I created a proof-of-concept project that reads tickets from Vikunja and creates PRs using Aider some weeks back.
Because it lacks specific criticism, your comment feels a lot like shallow dismissal template that can be deployed against any position one disagrees with, e.g. "This is what $(non-DIYers) tell themselves to feel better about letting $(electricians) have free reign with their money". Substitute with PaaS, banks, healthcare providers as needed.
My comment is shallow because it is not attempting to prove that active investing is better. It's saying, "I've heard people blurt out a list of reasons not to actively manage their portfolio and it's always been from a position of laziness or fear". There is a way to advocate for SPX or VOO or Fidelity 2045 and it's not by saying that these are the optimal strategies.
Several red-states in the US had already passed similar laws before the UK - see headlines where the pornographers responded by blocking visitors geolocated in those states because they figured there's no sensible way to verify age.
> Firefox is overall a fine browser. Still has 2% marketshare
Firefox used to have a 25-30% share before Mozilla shat the bed by neglecting it while treating Firefox like a money-piñata to fund a series of dead-end, copy-cat projects in their big-tech cosplay era.
Blaming Chrome for Firefox and Safari being shit (as reflected by the percentage of users who voluntarily use the respective browsers) removes their culpability. Chrome had to grow their share from 0%.
Chrome did that three ways: 1) performing better and crashing less than mainstream alternatives (just IE and Firefox, then) on non-Mac platforms (so, most desktop computers) for a good long while; 2) aggressive advertising to trick people who don't actually give a shit what browser they're running (or even know) into downloading it because "google said it would make my gmail work better" or "I dunno, google just told me to download this so I did"; and, later in the race, by 3) being the default on most Android installations.
One of three major factors involved actually being good, and I'd bet the other two factors overwhelm that one.
Many people have memory-holed Chrome's malware-grade tactics like including an installer in unrelated sourceforge downloads and now think that Chrome won strictly on its merits.
People have similar misconceptions about Google search.
The reality is that both had significant advantages over their competition when launched, but the company also used anticompetitive tactics to ensure dominance once those gaps closed.
I forgot how often individual webpages could take down your entire browser! On Chrome, it'd just crash the perpetrator tab. Chrome was also fast - really, really fast. Even if one was "tricked" into using it, you wouldn't want to go back to using other browsers - they put in the work.
> One of three major factors involved actually being good, and I'd bet the other two factors overwhelm that one
Counterpoint: Microsoft Edge on Windows has the same 2 factors going for it, but failing to replicate Chrome's ascendancy. Edge and Windows pleading with you to not install Chrome is kinda sad.
Edge (and basically every other browser besides Safari and Firefox) is a chromium fork though, so even though it only has like a 5% market share, it's still bumping up the engine's overall market share.
My point was that Chrome didn't win on marketing alone (as disproved by Edge dismal numbers). That said: browser marketshare metrics breakout the "brand" and not just the engine: for a long time, Chrome's rendering engine was downstream of Safari's WebKit before being forked outright as Blink.
That Microsoft abandoned Trident for a Chromium fork speaks volumes on the amount of innovation and engineering effort Google poured into Chrome/Chromium - I don't understand how it can be controversial to suggest that Chromium wins on its merits. The gaggle of browsers opted to fork Chromium rather than WebKit or Gecko because Chromium is best-in-class.
I agree that Chrome won mostly on merit, but I think it stays winning on inertia and marketing. There's just not that big of a difference now to the end user when using Safari or Firefox vs Chrome based browsers in my opinion. Safari's performance is fine. Let's not forget that Google retired their public benchmark suite because V8 wasn't beating JSC.
I can't say the same about the Trident based Edge though, it just wasn't as nice to use.
> The gaggle of browsers opted to fork Chromium rather than WebKit or Gecko because Chromium is best-in-class.
If I was making the decision to make browser, and finances were on the line, I'd pick Chromium just because other people have already done it. If I was doing it for fun, I'd rather just contribute to Ladybird.
> What is your argument that Safari is not a good browser?
Safari is often the hold-out on implementing features[1] that would be useful to users - presumably because it would make web apps viable on iOS, and compete with App store apps where Apple takes a 30% cut
> Using market forces to encourage more consolidation into a single engine is bad.
Competition on a level playing-field is not bad, even if you dislike the superior product (as determined by the free hand of the market.)
1. If memory serves: various APIs useful for PWAs were delayed or kneecapped on Safari
