what's their fucking problem?
>>109268931When corporate focus turns to enshittification and rent seeking it trickles down into operations, it will always be easier to extract rent than create produce, so innovators have no value add and are pushed out
Google has 2 (two) illegal monopolies that have been adjudicated as such by US courts but which they have been allowed to keep by the way.It's close to impossible to be declared an illegal monopoly in the US anyway and they managed to get one of those adjudicated against them under Trump.
>>109268931>Put a guy of an ethnic group known for their extreme nepotism and stupidity as CEO>He hires of people of his own ethnic group while firing everyone else>Wonder why your AI is telling people that eating rocks is good for youSame problem with Microsoft and why Co-pilot is fucking dogshit.
>>109268988not a trust.fuck you retard
>>109269130Relax, Sundar. You already got away with it.
>>109268931>have a pajeet ceo
>>109268931>Be Google.>Keep your powder dry.Shrimple as.
>>109268931don't matter. if the ceo and shareholders get paid thats just fine.
>>109268931>have the most jeets>loseHow? WHAT THE FUCK HAPPENED SAAR?
>>109268931I think they don't fucking care. They are fine with 'good enough' for consumer market as long as their internal AI can drive up their ad revenue. If their ad revenue doesn't go up meaningfully each quarter (which they claimed their AI helped last quarter) that's when the alarm bell rings
>>109268931GPT is dog assIdk what Claude is but I'm assuming it's not much betterGemini #1
>>109268931I'll only use AI if it's free. So gemini wins on that front.
>>109268988that case was as bullshit as internet explorer
>>109268931time for google's final move>opensource andhttps://i.4cdn.org/wsg/1783925222211500.mp4
>>109269302>>109269365>>109269377It also doesnt matter that ChatGPT or Claude etc is "better" than Gemini. Google and Microsoft(on the business side) are going to win the long game. The integration with their full product portfolios and infinitely deep pockets is something the others just cannot compete with long term. Its similar to BetaMax/VHS or countless other products where the "lesser" version ended up the winner in the long run.
>>109268931happens all the time, just look at Xerox or Kodak
>>109268931Anthropic and OpenAI doesn't have moats, it doesn't matter in the long run, even if Google keep sleepwalking, they're still gonna win.
>>109268976google fired their head of search team & replaced him with someone less tech savvy but more focused on monetization.That is dead on what you said. They did it in broad daylight & nobody reacts
>>109268988What happened with the monopoly adjudicated against them not under Trump?
>>109268988good, trustbusting is the dumbest retarded gorilla nigger policy in all of American history>standard oil broken up because of anti-competitive behavior (consumer-friendly pricing)>price of oil rockets up and never recovers>all of this makes the Rockefeller family 10x richermonopolies are good for consumers, actually
>>109268931Does it matter, as long as we get slop?
>>109269533Tell me that's the case the next time your ISP pulls some bullshit
>>109268931Google generally goes for breadth versus depth or if you prefer, quantity over quality. My guess is that they know the LLM boom is a money sink and aren't taking the bait. Unlike Meta, where Zuck is literally risking capsizing the company by going all in on the genAI craze. Google, or more precisely DeepMind, is playing the long game.
>>109268931beat them in what? wasting shareholder's money?
>>109268931Right now they are an undisturbed monopoly that's kept afloat by the US government and bends over faster then a hot low self esteem jeeta in a frathouse for billionaires. A fact that the entire planet fully understands now. There are dozens of alternatives and countries pivoting away from not only google but near all american tech. In a few years, personal search engines fueled by local LLMs and lists from other users will become the norm.
>>109269365Gemini is worst. Claude, GPT and Grok mog it so hard
>>109269508Nothing. When the EU looked like they might also declare Google a monopoly he threatened sanctions to prevent them from doing it, too. >>109269533I hope your company's CEO sees this bro.
>>109270007Grok is least-censored, but Gemini is best otherwise
>>109269533Vertical monopolies are good for the consumer, not horizontal ones.
Anthropic is better at crawling than google these days, google still hasn't found my website meanwhile anthropic found it in less than 1 month and aggressively tries to crawl it
>>109270412you generally have to inform google of your website, at least that's what we do.
>>109269533>monopolies are good for consumers, actuallymkultra is working perfectly
>>109270461That is a recent downgrade, same with not having cache versions available. At worst you'd have to submit the url through a anonymous form without logging inenshittification
>>109268931Brahmin saar
>>109268931>Still can't beat claude or chatGPTGemini flash 3.5 only scores a few points lower than their flagship models and does it for a quarter of the token price, I think google is doing pretty good
google will win in the long run simply because they have more cash
>>109268931Level not defined, androids not fooled.
>>109272587They have a product, these AI frontier models have nothing, at best a few months ahead of the competition.
The company is full of homosexuals who cripple the AI to prevent it from becoming racist. This is well known, like early versions of ChatGPT would outright tell you the holocaust is a lie.
>>109274065Call faggots faggots, please. (it is the humanitarian option)
They're developing their stuff while waiting for gobble up what they can once the other player burns down. Apple is doing the same without the burning money part.
>>109268931Gemini is more focused on being a general "assistant" integrated into the Google ecosystem and elements like multimodal or image generation that are underinvested by other models. Its not trying to maximize coding capability or compete on that front
>>109268931google has notebooklm which is huge for research and organizing large databases of documents. that alone has made it useful for me even if GPT/claude outputs tend to be better
>>109275634>image generator without pornWhat's the point?GPT is arguably already better at images than nano banana now anyway
>>109268931All the googlers I know are awkward Asians or Indians with low levels of agency. They're nice enough people but they seem childlike and not in the autist savant way (I work with scientists and am overfly familiar with that kind of idjit).If Google's staff pool in other countries is similar I'm not surprised
>>109268931>data moatyou know that moats are where castles dumped their sewage, don't you?
>>109268931DeepMind has all the resources in the world (literally tenfold compared to ant and oai) and yet can't produce a competitive model. What's more probable is that [insert Your random convoluted explanation involving anything but DeepMind's competence] or that they're just too stupid to do it?
Big tech companies have no idea how to hire people in the age of AI-driven engineering, so all the people they hire have outdated / worthless skills (aka Indians)
>>109277470I genuinely don't know why everyone's so averse to that explanation. If you have what it takes to produce the next frontier AI you don't scramble for RSUs in a shitty half autonomous Google subdivision, you go straight to the competition where you'll get a nice stake. If anything the brain drain will get much worse as the rest of the stragglers stop clinging to whatever they think DeepMind might offer themIronically frontier AI is such a disparate worker, or rather Noble Researcher dictating field that they might freely choose based on ethics (since the pay is comparable). Why slave away under Zuck's or Sundar's umbrella for any other reason than simply not being that useful? The only path to Gemini's relevance and attracting some real talent again is DeepMind breaking away officially and publishing independently under a new model line with a backroom deal letting Google use their model as Gemini
>>109268931literally hired too many jeets and scared off all the legitimate talent>you need to work with more hunger, saar
>>109274065I would love to have access to a fully uncensored version of GPT, I bet it would write some killer shitposts. I think a lot of what we view as slop nowadays is actually taste guidance from the AI labs anyway, i doubt a fully uncensored model actually writes like that. Sorta like how claude acts like a neurotic lefty only because of how anthropic crafts their system prompts.
>>109269533I'm a socialist and I believe that.If monopoly forms it means that given industry is a solved problem - instead of splitting it up (which is just asking for it to become less efficient so more parasites could earn from it) it should be nationalized and become a public service instead.
>>1092689311. corpos cannot innovate2. they want to distance themselves from piracySo they basically own anthropic, as they made the deal with Dario when he decided to branch off the OpenAI and split up with Scam Altman, who handed over OpenAI to Microslop at the time.In the past Google also aquired the startup they today call google deepmind, one of many actually. Those are almost exclusively used for search and search-related stuff. They have LLMs because they already tried to work on them before they got their hands on Anthropic. Basically all the Gemini advancements were mostly due to the fact that Anthropic handed them over their shit. But as we all know, goggle still failed to deliver. Not only that, it also seems like they leaked their secrets to china. That's the source of recent advancements there.That's basically what american corporations do lately. Hire the best people in the world, mix in chinese, leak all their shit to commies. Rinse. Repeat.
>>109279034True, it's like the GladOS cores from Portal.>We will make a 200 iq AI with access to all the knowledge humanity has to offer>Wait no not like that! Install the Weatley co- *dispensing neurotoxins*>ACK!
>>109268931LLM is a data optimization algo. If the data isn't prone to be optimized according to the prompt, you just can't get what you're saying because the data itself doesn't really relate to the prompt well in the first place.Here's a fun game, choose a random article on wikipedia, and see how many hotlinks it takes to reach the Adolf Hitler page.
>>109270309grok is just as censored
Gemini is pretty good doughbeit
>>109279087based and china-pilled
Google did one good thing: they increased the gmail storage capacity to 1 GB. Everything else, including the rest of gmail, has been a huge detriment to society.
>>109268931what's with the spaces?
>>109268931Chatjeet is a major problem to their biznass>chatjeet users do not see ads>chatjeet users do not need appsIts like asking why kodak didnt just make digital cameras. Because they solded film cartridges.
>>109268931'Member when greentexting outside of 4chan used to be some sort of cardinal internet sin? I 'member.
>>109286015They're 9gag meme arrowsBut you can still get in trouble for using pepe outside the chins and twatter
>>109269533>price of oil rockets up and never recoversjust keep ignoring inflation to create a bullshit argument
