Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Google I/O 2023

io.google

Post reply on HN
Print view View original

431–440 of 443 posts

Re: Google I/O 2023

#431

Post by fallat »

saltcured wrote: post #294
throwaway894345 wrote: post #216

Earlier quoted context omitted.

I think it's a lot less confusing when you understand that "hacker" doesn't describe a kind of politics, but rather connotes an interest in technology.

Depending on who you talk to, hacking doesn't even mean technology. It's more of an artful blend of craftsmanship and trickery in whatever problem domain you face. But I think the confusing aspect of HN and the VC influence is the number of folks here who happily conflate technology, business, entrepreneurship, and capitalism. If you see these as separable topics, you find that a lot of the confused threads here are…

> Depending on who you talk to, hacking doesn't even mean technology. It's more of an artful blend of craftsmanship and trickery in whatever problem domain you face.

Exactly this. I use the term hacker in its original form.

Re: Google I/O 2023

#432

Post by fallat »

esafak wrote: post #373
smoldesu wrote: post #362

Earlier quoted context omitted.

And when Jobs got the contract from Atari, he lied to Woz about how much it paid so he could take the lion's share. Fraud culture is unfortunately part of their blood too, something we ought to excise if we want the tech industry to be taken seriously in the long-run.

Everything I read leads me to believe that Steve Jobs was not a virtuous person, though I keep trying to find the silver lining. How should one test for ethics? A reference check comes to mind, but it is not a test that you can apply fairly.

> How should one test for ethics?

You ask if that's how you'd like to be treated.

Re: Google I/O 2023

#433

Post by phyllistine »

ducttapecrown wrote: post #329
noisy_boy wrote: post #278

To all the people who have been espousing how they appreciate cover letters, good luck - you'll be swimming in them, all AI generated. There will be a sea of all good, all positive fake text and text will lose its meaning. We used to appreciate a nicely written mail - now AI will be writing perfectly composed letters. We will learn to appreciate a badly written mail.

Texters and readers of Homestuck have been practicing for decades at appreciating badly written mail. I believe human-to-human knowledge transfers will still be fine :). I'm fucking scared of the incoming spam tsunami though!

>readers of Homestuck

I will develop a typing quirk for the express purpose of ensuring that I wrote the email instead of AI

Re: Google I/O 2023

#434

Post by ricktdotorg »

intrasight wrote: post #15
danwee wrote: post #2

I was very enthusiastic about Google I/O in the past (around the year 2010). Full of nice tech talks; I learned a lot. These years, I couldn't care less. It's all business talk. I guess developers are not the target audience anymore (and for newcomers, perhaps the "I/O" may be misleading).

Same. Google invited me to participate and have a booth - in 2010 or 2011 - because we were an early business adopter of their APIs. It was a cool event. I think the hot new tech that year was NaCl. I don't think startups get invited now just for using new Google tech. And is a reminder that this year's hot new tech will be forgotten in a few years.

> I don't think startups get invited now just for using new Google tech

You're right about I/O. The once-fat "invite adopters to conference" coffers seem to have been transferred to their Cloud Next conferences instead.

Re: Google I/O 2023

#435

Post by danudey »

sebzim4500 wrote: post #343
RandomThrow321 wrote: post #306

Maybe going against the grain here, but this was the best Google I/O I've seen in a long time. In the past, they seemed liked incremental updates on their hardware (which I don't care too much about) or some new products like the pixel watch / glasses (which I probably won't use). I knew there would be a huge focus on AI this year, but I was pleasantly surprised by many of the new features and how quickly these integ…

I am currently in a superposition of being very excited and very underwhelmed. The wavefunction will collapse once I can actually play with the technology and find out if it is anywhere close to as good as they claim.

It rubs me the wrong way that Google fired their whole AI ethics team and then got on stage and talked about how AI can be bad for people and we need to guard against people using it unethically.

They've realized that, at this point, they may well have lost the AI chatbot war, and now they're pivoting to "we can keep these other, rogue LLMs from being as bad as ours would have been if we'd gotten there first".

Re: Google I/O 2023

#436

Post by computing »

ShamelessC wrote: post #393
computing wrote: post #198

In 2018 an obnoxious researcher working at a company known at the time as DeepMind told me (who was working in healthtech) that "AI in healthcare is a solved problem". It's 2023 now. AI in healthcare is a rounding error and will likely stay that way for a decade or longer. Google blew a 7 year lead of being an AI-first company to a non-profit. DeepMind is now a team within Google, similar to Ads, Drive, and Shopping…

I’m curious what they had been referring to at the time? Was AlphaFold out then? Or were they just clueless?

image labeling, although even then they had some oopsies with classification in Google Photos

Re: Google I/O 2023

#437

Post by CydeWeys »

r0m4n0 wrote: post #360
CydeWeys wrote: post #354

Earlier quoted context omitted.

> Final Cut Pro is a thing people already use in their workflow, becoming more convenient for them. People have depended on this as part of their production-ready workflow for over 20+ years now, and FCP-folks seem to have a lot of trust in the application and related tooling, that Apple isn't going to leave or abandon Final Cut anytime soon. All of that may be well and good, and yet I can't remotely get excited by a…

My theory is there are a lot of folks that needlessly bought expensive ipads that sit around and collect dust when their kids aren't doodling on them. Finding out that Apple is bringing an app that could bring potential productivity to a bad investment got a lot of people excited. People will soon learn that another app was not the reason why they aren't using them.

I have an iPad sitting around collecting dust (literally), but I'm not about to pull it out to use for something that works better on a desktop.

Re: Google I/O 2023

#438

Post by cubefox »

gabereiser wrote: post #285
cubefox wrote: post #224

Earlier quoted context omitted.

As you wrote this, they announced PaLM 2.

Yeah, and I realize my point of view is unpopular here. Integrating across all their services is just doubling down on what they already took from GPT. I didn’t see anything that wowed me. Cool, so I can LLM in Gmail and workspaces. Cool, so I can build a job description based on what others are hiring for instead of my company’s needs. I’m not saying that any of this is easy, just that it’s simply novelty. PaLM 2, w…

Yeah, but it seems their real "big new model" is Gemini, not PaLM 2:

> We recently brought these two teams together into a single unit, Google DeepMind. Using the computational resources of Google, they’re focused on building more capable systems, safely and responsibly.

> This includes our next-generation foundation model, Gemini, which is still in training. Gemini was created from the ground up to be multimodal, highly efficient at tool and API integrations and built to enable future innovations, like memory and planning. While still early, we’re already seeing impressive multimodal capabilities not seen in prior models.

> Once fine-tuned and rigorously tested for safety, Gemini will be available at various sizes and capabilities, just like PaLM 2.

It sounds like this is the first "joint project" of the old DeepMind/Brain groups since the merger.

Re: Google I/O 2023

#440

Post by chaxor »

int_19h wrote: post #376
sebzim4500 wrote: post #366

Earlier quoted context omitted.

>but the migration of Bard to PaLM 2 it will exceed whatever OpenAI currently has I really doubt that. If Google had anything that could seriously compete with GPT-4 they would be shouting about it from the mountains. Not publishing a comparison where they only include GPT-4 in the two benchmarks that they win.

Here's my entirely non-scientific comparison: https://imgur.com/a/HWLgu3c If that is really the best that Google can do after working for several months on one-upping GPT-4, that's just sad, and doesn't bode well for them.

Google in some ways is far better than OpenAI in their AI capabilities, but it mostly comes down to the fact that they weren't making products, when OpenAI was.

Deepmind has probably the best research in the AI area fundamentally, and the capabilities they have shown academically are much better than any other group. However, they have been seemingly completely disinterested in making any product usable for the public in the area - which takes a good deal of engineering and glue. So it will take a bit of time for them to make that happen. For example, OpenAI was not the first to use RLHF to train a language model (Sparrow from Deepmind was the first to conceive of this idea). Additionally, Deepmind has far better people in the general GNN and algorithmic reasoning area, which is going to shape up to be far more important in the coming years. That's of course not to say OpenAI doesn't have excellent research or people (Sutskyver is integral to an incredible number of monumental achievements in NNs), but the company as a whole focuses on product far more than fundamental academic research, and that has a potential to creep up on them.

Obviously Bard from the scientific side seems like it was probably hastily thrown together in an afternoon by one guy, and then sent to the marketing/we dev/etc group to put it public facing.

Post reply on HN