imimayj1337

And not just hyperscalers... even the neoclouds who profess to have fantastic availability and GPUS NOW just tell you to jog on. Though of course they do that after they've signed you up and promised preemptible instances :)

why on EARTH would I contribute to the effective doxxing of someone because of 'reasons' nebulously alleged in a relatively niche community and published to HN????

It does seem pretty insane to me that only on max setting (and after waiting 27 seconds) did you actually manage to receive an SVG where it appears that the pelican's legs are straddling the bike correctly.

Edited to make clear what it looks like to me and what I was looking for.

you searched simon glass u boot and nothing came up?

I think the wine test was actually the part which blew me away the most. And they say it outperformed humans (albeit don't know if they were wine 'experts') on 19/20 tests?! Crazy work

I also found it refreshing - particularly given the semi lobotomised LLM slop we're seeing most articles written in nowadays. It's nice to see someone's actual tone coming through even if it is peppered with profanity!

I'm... really uncertain on their attitude towards naming and shaming him. It's quite easy to find information relating to him personally online due to what they have posted, and yet the allegations made against him seem nebulous at best?

I didn't realise this was essentially a plug for a service until I got to the end. I do agree that it seems we are sinking more and more time into chat interfaces (and now all of my google dooming has been superseded by a demented chatbot which routinely provides me with incorrect or hyperbolic information so that joy has been sucked away). I like how they have all of the resources they used for research. I'm still not certain whether this service itself actually diverts away from attention mining however (and in that case, the article itself though beautifully written is slightly disingenuous)

ndas aren't legally binding in the UK. what they have signed is non-disparagement agreements (not allowed to publicly disparage the company)

if you read the article it actually outlines that this is happening in the UK, where whistleblowers are legally protected if what they choose to blow the whistle on is _in the public interest_ so actually this comment does seem pretty spot on: They don't have anything beyond moral disagreements on what the company is working on (the article mentions military applications of AI), and are thus taking a moralistic standpoint on applications which are otherwise within the laws of the UK. Otherwise they would be able to (and would legally be protected) if they spoke out.

I know we just had the carbon announcement which was distinctly confusing, but comparatively I think that this is an exceptionally well-written article. I do believe that Google has some exceptional scientific minds as illustrated by their ability to take highly technical concepts and explain them in an understandable way for the layperson.

As for the technical nature of this, I'm not sure if I've totally understood correctly despite how well it's written, but are they essentially holding back sort of unweighted parameters built in to the model architecture which they can then populate with new context during 'sleep' and then using the techniques outlined in the 'dream' section to reinforce what gets stored in those MoE weights? That to me seems like a solution which only Google is best placed to solve due to their fully integrated AI stack and historic experience with hardware optimisation, but which would be rather hard to do if you are a more resource and compute constrained organisation

It's true! It feels like we've been talking about harness optimisation and 'cool features' available in the cli tools for months at this point, but ostensibly there has not really been any significant upgrades to these harnesses since at least Claude Code imo. It does feel like a contrived way to harvest more and more information and test each conversation/action tool, to the detriment of those of us actually using them!