I mean... there is Boeing, which is arguably more horrifying.
With AI it seems to be internal testing produces wild results, which is unsurprising to me, all of these models can be jailbroken, and as long as that's the case, they're all likely to be exploited maliciously.
Not just internal, Anthropic’s S-1 literally covers the risk of ending humanity. They are telling their investors a public company could eradicate humanity. Give us money, the risk you have to evaluate in your financial framework is that we could eliminate humankind.
I’m repeating it because that has to be internalized: the largest IPO ever will be for a company that casually consider they might kill us all and their own employees too. Including their children.
> Of course that not everyone keeps the same email address
Or Microsoft decides to nuke your hotmail account from the 90s and someone else takes the name from you, along with all your old accounts that you are requesting for a password for. They really should have evaluated a little better how they determined which "inactive" accounts were actually worth nuking.
Every time I hear stuff like this, I think of that Office Space thing... you don't want the Engineers talking directly to the customers.. well it sounds like the Engineers are also handling all releases and business decisions willy nilly.
Because I have worked with responsible engineers, but it seems like Google is overrun by them there's a lot of really bad business stuff going on. It's like when I think of AWS, nothing on there is named for business people, its all really bad names that IT / Devs come up with.
AWS as a whole is sold to businesspeople, sure, but the individual AWS products are definitely being sold to engineers. S3 and EC2 don't need to have layperson-friendly names, since laypeople don't understand what they are or why one would or wouldn't use them - those are CTO/tech lead decisions.
I still have simpler conversations when talking about Azure services and naming them, they have reasonable names. I never did anything with GCP so I cannot compare, but I assume their naming scheme is also reasonable. It feels to me like AWS was basically they decided to offer some of their internal things externally, nobody renamed anything, and then they just ran with it.
> In theory MS has the same opportunity (plus they have GitHub so, you know, dev eco system too) but seem be blowing the strategy.
Yeah, they have Phi but offer it nowhere on CoPilot as far as I know, I can't even register for copilot, which is bizarre. They came out with "MAI" but... nobodys talked about it since, not sure if its even used by anyone? They're as bad as Mark Zuckerberg is about it.
I do appreciate both Microsoft and Google for releasing small models, unlike Anthropic and (not so) OpenAI.
Yes and no. Microsoft is likely looking at open weight models and going oh, this is just as good or better than we could do, without the training costs, and we simply are an inference commodity provider where azure already operates. Count us in.
The thing is, Google doesn't have to scramble and freak out every time the others do some impressive update, because they have stable income, the other two do not.
Kind of and kind of not. Their main resource of serving ads is getting legitimately eaten away. The route to getting AI to replace those ad dollars isn't clear and the capex to build it out is significant. Google has to manage a variety of assets that are currently lucrative but under threat. Yes they don’t have all their chips in one basket but it isn't clear that helps them in the race… it hasn’t so far.
their stable income is eaten away by the quickly rising income of the labs, they will be out of the race once the labs reach a certain threshold of income, rather sooner than later. Unless many people install Google surveillance anti-gravity thing on their system which I doubt
I've been lowkey rooting for them, not a fan of Altman, and I don't know if I trust Anthropic either, I'm not Google's #1 fan, but I think they can definitely innovate heavily in this field, they have a lot of key things as has been mentioned.
I'd add that we've seen how the leadership/reputation of different AI leaders, and with 2 decades of history with Larry and Sergey, they seem to come up on top as the most respectable.
Did they ever really abuse the power Google could have wielded?
I could be missing something, but for the most part they seem to just get down to building and pushing technology/science forward and avoid drama rather than welcome it.
> Did they ever really abuse the power Google could have wielded?
Plenty of times. Perhaps the worse is pushing Manifest v3 with no support for real ad blockers in Chrome. Not to mention no support for ad blockers in Chrome mobile.
Also, their stealing of newspaper content.
Their new push for AI summaries instead of actual search in their search product, stealing even more clicks from sites that actually produce information.
YouTube pushing Shorts to everyone, and now the weird text posts that you can't hide.
In general their gigantic push for ads everywhere all of the time.
> Did they ever really abuse the power Google could have wielded?
The "Do no evil" never being an official slogan and people saying it's no longer true for the last 10 years or so, they're just really massive. They do mess up a lot, but when you're Microsoft, Apple or Google sized, its nearly impossible to get everything right all at once.
Google, with the help of Facebook, destroyed the independent online publishing business. They made it impossible to maintain an honest publication and tunneled users to low quality websites until there's not much left of respect.
There are pictures of Sergey in the files, but they are rather milquetoast. He's just sitting next to David Brooks of the NYTimes. I don't think there's any evidence he's engaged in sex abuse.
Ehh, the internet was worse before Google. Google actually became popular for having a nice blank page with only the search bar. Even the ads they had later were much cleaner compared to yahoo, altavista, lycos, aol and other firms of that time.
I am not rooting for them, I don’t own stock from either just that like GP I was sure by now Google would have eaten others’ lunch. I was expecting a Gemini 4 before Mythos/Fable and Astra just from what GP pointed out: a cash hose, data centers, good engineers. But I underestimated the investment others would keeping absorbing so started to lose confidence in my Google prediction. Let’s see how Gemini 4 holds up. Are we going to see a flurry of scientific and mathematical breakthroughs from it?
Please check Gemini's t and c, as the last time I checked when they had a good pro model (3.1) you couldn't turn off training on your data with a paid account .
Even connections won't save you anymore. I have an account with over a hundred connections but now they want to see a government ID for whatever reason. Won't let me do anything on the website until I do.
> 1. Collaboration between always-on agents is a really, really powerful thing. It allows for domain-specific expertise that doesn't overload the context window, while still allowing for access to knowledge if they need it.
So why don't I just have Claude write notes and summaries on specific things, and then it will always have a knowledge base? Am I missing something? I mean, if Dots and Grok Bot are not an extra charge / extra compute, then I guess that's fine, but in Claude Code you can make a AGENTS.md or CLAUDE.md file, and if you put it into any directory, Claude will read it when accessing that specific directory, so if its in your root where you launch Claude, your new Claude instance will read it and have all that in a dedicated smaller context window. But as it edits code, it can read the smaller ones too.
With AI it seems to be internal testing produces wild results, which is unsurprising to me, all of these models can be jailbroken, and as long as that's the case, they're all likely to be exploited maliciously.
reply