jfoster

47 ms·

631.
by jfoster 4y ago
Because part of being big & successful if you're Google involves operating a business. If your approach to everything is "why risk it?" then you eventually just choose to shut everything down.
632.
by jfoster 4y ago
The spam that is getting through seems like really obvious spam. Not just for a human, but for a classifier, too. I looked at an "out of the blue" example just now: 1. Sent from a Gmail address where the name doesn't come clo
633.
by jfoster 4y ago
> Would that be harmful to Google to have Google Sheets be used to distribute? Yes. If it's a private file (not shared), the question doesn't even make sense. Some degree of control might make sense for 100% publicly shared con
634.
by jfoster 4y ago
Someone must be working on one, in light of stuff like this, surely?
635.
by jfoster 4y ago
It was a solution. Google Drive was great until they turned user/customer hostile. Arguably, no one could really have seen this coming, because this doesn't even seem to make business sense. Google are seemingly acting against the
636.
by jfoster 4y ago
> Maybe the flagging algorithm looks at private sheets with much more suspicion. You could be right, but jeez... shouldn't it be the opposite?
637.
by jfoster 4y ago
It's a shame that Google are choosing to focus on policing what people think of as their private data. Meanwhile, they seem to be taking their eye off the ball when it comes to the spam filtering that many of their users might wish for
638.
by jfoster 4y ago
Yeah, this is key. Improving a product in the direction of customer intent vs against customer intent.
639.
by jfoster 4y ago
This isn't a problem because a single bad review doesn't outweigh many good reviews. Something I think the next generation of reviews should take into consideration, though: places change. A place that used to be good can take a t
640.
by jfoster 4y ago
None of that changes that it does seem to be the key criteria that the outcome of case will depend upon.
641.
by jfoster 4y ago
It matters because that will probably be one of the deciding factors should this eventually be assessed in court.
642.
by jfoster 4y ago
When it comes to theft or murder there are broad laws. There isn't one law to say that you can't murder using a gun, another to say that you can't murder using a knife, and so on. In this case, it seems there's a less br
643.
by jfoster 4y ago
Isn't it silly that there needs to be laws written to ensure consumers simply get what they paid for?
644.
by jfoster 4y ago
The thing about spam is that as it becomes more evasive, it's also less clear for the humans it was intended for. Once they raise the bar high enough, the spam will be gone.
645.
by jfoster 4y ago
It sounds like you're assuming that Google sticks to a single approach and doesn't waver from it, despite spam still being present. They wouldn't do it like that, though, would they?
646.
by jfoster 4y ago
I think autonomous cars are something that must happen, and the sooner the better, but I am also a bit concerned that they could represent the ultimate weapon if a malicious actor gained control over them. One battery fire isn't a prob
647.
by jfoster 4y ago
Even humans can't safely drive in all sets of circumstances. There may even be circumstances where autonomous vehicles can drive, but humans can't or shouldn't.
648.
by jfoster 4y ago
Perhaps someone who understands the ad industry can help me understand it a bit better. Does Google ban forums because there's simply too much inventory around otherwise? I realise the stated reason will be that it can be unpredictable
649.
by jfoster 4y ago
It's no secret that free Gmail has noticeably failed to catch a lot of obvious spam recently. Seems plausible that they might be preparing for the introduction of a premium spam filter, because there's not too many other explanati
650.
by jfoster 4y ago
I don't think they're losing the battle. Many of the spam emails that Gmail misses don't require a fancy AI to detect. What I've noticed is that even simple heuristics could automatically catch them. As best I can tell,
651.
by jfoster 4y ago
The DALL-E mini tool is probably the first thing like this that is open in that way. I'm not 100% sure, but I don't think it has limitations on how you're allowed to use it?
652.
by jfoster 4y ago
I didn't really think of it as having full access between two brains, but more as having a non-verbal bridge between the two, allowing for more direct sharing of individual thoughts, emotions, etc. Of course, if it's just a wirele
653.
by jfoster 4y ago
If you're going that far, I guess you might as well set up a commercial search engine or SEO tools business off the back of it. At which point, the thinking may shift away from open source toward something proprietary, unless there wer
654.
by jfoster 4y ago
For most software, there is an open source attempt aimed at producing something more or less equivalent. Is there any such thing trying to attempt the same level of scalability as Google or Bing? I'm aware of Apache Lucene, Elasticsear
655.
by jfoster 4y ago
Neuralink is going to be so interesting if it can allow two brains to communicate more directly than words.
656.
by jfoster 4y ago
I think there's currently some barriers in doing that: 1. OpenAI claim ownership of the content produced. 2. The content policy forbids what you're describing. (commercial use is ruled out) Content policy: https://labs.
657.
by jfoster 4y ago
You don't think online businesses will treat a customer who enters a different date of birth each time as less trustworthy? You don't think BigTech ever cancel accounts being used by regular people? You don't think companies
658.
by jfoster 4y ago
> As a small revenge I always use an absurdly old birth date on these forms. This way if they are collecting this data, my input will be discarded or pollute their datasets. Or the micro-frauds get detected, linked to your credit card &a
659.
by jfoster 4y ago
That's pretty much my point. We have very few tests for this. In the absence of tests, when people express an opinion on whether something is/isn't sentient, what are they basing that on?
660.
by jfoster 4y ago
The Turing test is imperfect & LaMDA is probably not sentient, sure, but what's a better test?

More ›