> how is the oldest model, the Pro model?
Because apparently their Pro models had issues.
Wouldn't you rather a company be honest about that than hiding it from you?
The other end of the scale is Microsoft, where everything is Copilot.
4,608 karma · joined August 10, 2016
Because apparently their Pro models had issues.
Wouldn't you rather a company be honest about that than hiding it from you?
The other end of the scale is Microsoft, where everything is Copilot.
I started using AWS when EC2 and S3 first came out, but was forced to switch to Gcloud by the most recent company I joined. I was pleasantly surprised. In many ways it's more techie-focused than AWS, which is more enterprise-focused. And their managed Kubernetes is second to none.
I have an AI Pro account - which I decided was a no-brainer because of the 5TB of Gmail / GDrive / GPhotos space it gives - and this all just comes along with that.
One thing to check, though, is that the client is running the latest version. At one point my Antigravity GUI, which had been working fine, started listing only Claude and GPT-OSS models. It turned out to be because the GUI needed to be updated. I guess the Google API endpoint had changed behavior without any backward compatibility.
So yeah, they're hilariously bad at rollouts, but it's not that difficult if you're technically inclined or ask an LLM.
> Instead, labor and computing power are being diverted toward a singular “frontier-model effort”
It’s a consolidation, not a withdrawal.
This isn't a case where they've identified a suspect using some specific investigative lead, reasonable suspicion, or probable cause.
Instead, they're using a widespread fishing tactic - a "dragnet" - to gather information in order to generate suspects without any reasonable suspicion. This kind of behavior has been found illegal in the past, because it clearly violates constitutional protections.
Sure, they can spend money on R&D, but then they’re competing with every other new venture in the world - the only benefit they have is money, but collectively, everyone else has more money than they do.
This is why when private equity takes over a mature business, they almost always shut down any R&D the company is doing. Silly mature business, that’s not the reason for your existence!
Many of those data points are contaminated by a long tradition of sex- or gender-segregated races.
Here are a couple of interesting studies, which I assume are the ones the SciAm article is referring to, but I haven't checked.
https://pubmed.ncbi.nlm.nih.gov/38684111/ :
> "There is a prominent sex-based difference in athletic performance such that males outperform females by 7%-14% in races from 100 m to marathon. In ultramarathons, the difference is often much smaller, leading to speculation that females are "built" for the sport. However, data are confounded by the low number of female participants; just 10%-30% in any given race. This study compared data from two ultramarathons where males and females competed in comparable numbers. [...] In conclusion, the sex-based performance discrepancy shrinks to 1%-3% in ultramarathons when males and females compete in comparable numbers."
A key point in the above quote is "data are confounded by the low number of female participants; just 10%-30% in any given race." This calls the "many, many data points" that you mentioned into question.
Another study, https://pmc.ncbi.nlm.nih.gov/articles/PMC12068597/, of 56 suitable races, found that "pacing was not consistently affected by sex or age."
Re keeping an open mind, the corollary is that if you're not willing to ever question your assumptions - a.k.a. biases - you're not going to be able to update your understanding based on the facts.
If you really evaluate the best options, it’s very likely that a Jev-style decision model won’t give the best results.
Laplace expressed this in "A Philosophical Essay on Probabilities" in 1814: "We ought then to regard the present state of the universe as the effect of its anterior state and as the cause of that which is to follow. [...] For [an intelligence that has sufficient information] nothing would be uncertain, and the future, as the past, would be present to its eyes."
Superdeterminism essentially extends this idea of deterministic physical evolution to quantum mechanics. It could be considered an appropriate default simply because it was previously considered to be, not only the default, but the strongly believed truth.
> Superdeterminism says that there’s no way to pick your experimental choices in a way that reveals things because any possible random number sources you could use have contrived correlations that prevent that
You're arguing against a consequence of superdeterminism, not its motivation. Its motivation is as I've described.
The word "contrived" is doing some heavy lifting in your statement. Nothing needs to be contrived in superdeterminism, even if it seems that way to an observer within the universe.
The explanatory problem you've alluded to only arises because we're trying to understand the universe from within the universe. The universe is under no obligation to make itself sensible to us.
Essentially, you're saying little more than that you don't like the idea of superdeterminism.
Human output is not reproducible either. I take it your argument is that we should remove humans from the loop, then?
No? In that case, you have some reconciliation to do.
Whether or not that's true, you have to keep in mind that it's not true of everyone.
I'm a passionate advocate of LLMs. I implemented my first neural net, for image recognition, in the late 1980s (based on Bart Kosko's work, I'm not claiming to have invented anything.) Now, I use LLMs daily for coding and other tasks. I don't usually "vibe code" unless it's a simple enough project that the code quality really doesn't matter - a simple example would be data integration, where as long as the data in the target system matches the data in the source system, I don't care much about how the sausage in between is made.
But otherwise, I'm approving or denying every tool call, like a caveman, and when they mess up, I explain to them why. I interrupt them and redirect them when they need it. I change code myself and let them know. With all that, they're still incredibly useful - more useful than most "junior programmers", realistically. No human software developer can hope to match LLMs on many significant dimensions of development. But that doesn't mean experienced and competent humans have no role, at least currently.
The disconnect here is that your characterization of "the low quality of their output" is too simplistic. The quality of their natural language writing is different, for the kinds of reasons stated in the OP. Some of the accusations of "lower quality" really boil down to "too compressed for humans to easily understand", again as the OP points out. And the quality of their code has a lot to do with how they're prompted.
One of the biggest barriers to good LLM output - both writing and code generation - is that they're trained on data generated by humans who fit George Carlin's observation that half of them are stupider than average. When it comes to coding, on average they don't have enough education not to think that imperative code with minimally-restricted mutation all over the place is a good idea. They think dynamic typing, so that code can't be statically verified, is a good idea. They reflect our own collective limitations. If you recognize that and redirect them to sources of good practices, they can be much better.
In short, they're much more of a mirror than most people realize or admit.
With all this noise, it should not be forgotten that we appear to have figured out roughly how a significant component of intelligence works, and have real life, useful proofs of concept. All the technical criticism, to me, is just coming from people who don't seem to understand how intellectual advancement works. The political criticism is another matter.
That makes sense, because fluent readers are using more System 1 processing to read - they’re not thinking about it for the most part, they’re just pattern matching instinctively based on text they’ve been exposed to previously (“trained on”). AI writing breaks those patterns so compromises their fluency.
Whereas a second language reader or poor reader is exerting more conscious attention to understand the text, so it doesn’t matter as much if it doesn’t conform to standard patterns.
They’re difficult for humans because if a sequence of words hasn’t previously been seen, you have to expend more effort to make sense of it.
Something like “customer onboarding workflow automation" might qualify - it makes sense, but it’s not a commonly used phrase. AIs use a lot of these jargony sequences.
Similarly, superdeterminism is simply the application of the determinism that was believed to apply to physics before QM, to QM itself. In a way, one might have expected it to be the default interpretation.
Superdeterminism and retrocausality both allow for local hidden variables - it’s just that they don’t satisfy Bell’s assumptions.
There are reasons to do this, like support for minimal environments, consistent cross-platform behavior, simpler deployment, and application control over root certs.
“Authentication” is extremely far from being the only relevant characteristic of security. That in fact is the whole problem with security: it’s a pervasive concern that has to be considered throughout a system, at every level.
Transformer models can transform natural language to another target language.
That’s the commonality. What’s the distinction you want to make?
As it stands, the headline is misleading and clickbaity.
> the "in mice" doesn't matter
Of course it matters, because that has a direct bearing on how applicable it is to humans and how much time it’s likely to take to be of any benefit to humans, if it ever is. That’s the whole point being discussed.
Also see this comment to help understand why “in mice” is a huge qualifier in this case, because mice don’t naturally develop Alzheimer’s: https://news.ycombinator.com/item?id=50218606
They seem equally bad at both.