Hacker News .hnnew | past | comments | ask | show | jobs | submit | manc_lad's commentslogin

they're going to start training a pelican riding a bike specifically on these models soon. it's the key global benchmark!

They will have a long time ago. By now Simon's meme will be well represented in training sets.

This summarises my thoughts as well. People used to give away children that were too much of a burden. With higher mortality, some were also not expected to make it.

We're in a societal place where we have set the bar high in terms of an expected level of education and quality of life for our kids. kids are expensive and we've grown the population massively. There is also a social stigma associated with having lots of kids in Western countries.

I worry about my kids. But im always fascinated when they stretch the boundaries and show me how resilient they are. So I let them push limits but explain the pros and cons hoping they build their own feedback loops with some sense of perspective.

It's a delicate balance as a parent. I'm consistently fascinated how others parent. It's amazing how changes in parental style can be generational and show how long the changes will take to change.


> People used to give away children that were too much of a burden. With higher mortality, some were also not expected to make it.

Don't confuse that for them not loving their children or not being devastated by the deaths when they happened. We don't know how often kids were given away, but there isn't strong evidence it happened often.


Zelda Breath of the Wild attempts to approach this with an interesting interface.


Tears of the Kingdom?


It seems more and more like an inevitability we will run models locally. Exciting and concerning implications.

If anyone has any suggestions of people thinking about this space they respect, I'd love to listen to more ideas and thoughts on the developments.


I think the main limitation, right now, is hardware. For GPUs the main limit is the VRAM available on consumer models. CPUs have plenty of memory but don't have the bandwidth or vector compute power for LLMs. This is why I think the Strix Halo is so exciting: it has bandwidth + compute power plus a lot of memory. It's not quite where it needs to be to replace a dedicated GPU, but in a few iterations it could be.

I'm interested in other opinions. I'm no expert on this stuff.


How does the shared memory model for GPUs on Apple Silicon factor into this? These are technically consumer grade and not very expensive, but they can offer a huge amount of memory since all the memory is shared between CPU and GPU, even a midtier machine can easily have 100 GB of GPU memory.


If you squint the M4 is the same as the Strix Halo. The M4 has roughly

* double the bandwidth;

* half the compute; and

* double the price for comparable memory (128GB)

compared to the Strix Halo.

I'm more interested in the AMD chips because of cost plus, while I have an Apple laptop, I do most of my work on a Linux desktop. So a killer AMD chip works better for me. If you don't mind paying the Apple tax then a Mac is a viable option. I'm not sure on the software side of LLMs on Apple Silicon but I cannot imagine it's unusable.

An example of desktop with the Strix Halo is the Framework desktop (AI Max+ 395 is the marketing name for the Strix Halo chip with the most juice): https://frame.work/gb/en/products/desktop-diy-amd-aimax300/c...


I am also very interested in AMD's Strix Halo for running LLMs locally. For that I have a Framework Desktop in order (batch 1!). Alex Ziskind on Youtube does videos comparing Strix Halo, M4 Mac mini and MacBook Pro, Nvidia 5090, etc. including power consumption. The only downside is one has to pull out the numbers from the videos, there's no tables or anything. Here is the recent video with testing Strix Halo and a Mac mini: https://www.youtube.com/watch?v=B7GDr-VFuEo


Apple has machines with 2x and about 3x the Strix Halo bandwidth by doubling up the memory buses. These get expensive though.


Honest question: what is the concerning aspect to it?


I don't know, what's worst with people running LLM locally compared to running any software locally?

There is nothing fundamentally new in having freedom in edge of societies. Yes it can lead to horrible situation, like someone kill neighbors, using the single handable bright new tool available to all. But that's far less of a concern than having the powerful new tool staying in full concentrated control of the greediest humans out there, who will gladly escalate any hindrance to genocide whenever something doesn't fit their perspective.


it was my experience of college. many I know would agree, and few would agree with you. I'm sure there are some that didn't feel this way, but strange sweeping statement to make.


I agree with this. If she's adding value and her input is worth her place on the cap table then it should be easy to resolve.

You do need to think carefully about what a breakup would look like. Relationships can go sour and if that happened what would the arrangement be if they couldn't work together.

Also, who makes the decisions? Does her allocation give them both a majority of voting rights?

It's not a typical arrangement, that doesn't mean it can't work but your co-founder has added complexity into the process.


Burnley Football Club | Product Manager | Burnley Lancashire UK | full time | on site

Do you want to help build the football club of the future? With big ambitions to reinvent how a football club is run, I'm looking to hire a Product Manager in a new technology department to work in every area of the football club. HR, medical, retail, stadium hardware and more. It's a full stack technology job. Initially building partnerships and then custom tech it's exposure from the ground up in a startup like environment.

If you're interested in complex and high performance businesses and working in focused commercial businesses then this is a big challenge.

Apply here https://careers.burnleyfootballclub.com/job/523819

Or reply to this comment and I'll answer any questions you have.


Can I submit my Football Manager records with my CV?


Very interesting to see this in HN. Dream job for someone who loves football and product building. Hope you find someone great


This is great to see. I'm no longer local but I imagine for some this would be a double-whammy dream job. Good luck finding your person!


Thanks. I joined as CTO at the start of the year and it's a fascinating business. really pro-active board and big ambitions. Plus free tickets to games! What's not to love.


This sounds really fun, do you have any pay indications in mind?


multiple ways to make that work. LTIPs, share options, direct equity in subsidiary etc


MSFT comp is shit though

OAI comp was high based on equity and its crazy valuations


For us maybe, but they have pulled off some high profile hires in the past… Brendan Burns (one of the main k8s guys) for example.


Microsoft comp is actually not bad at higher levels, which I assume will be given to all OAI people that will join.


Is Microsoft compensation for top AI talent also bad?


Received wisdom has been “not competitive”. I wonder how the MSR folk feel about all of this, too.


OpenAI never gave equity.


Using the same logic, you could say coal is renewable depending on your time horizon, no?


No, fungi evolved that can digest lignin now ;).

On a less facetious note, Solar needs time to renew the energy it made the day before as well. Time scale is very important to questions of renewables, and people have been sustainably burning wood for a long time. Unsustainably too, but I would bet that’s not the case here if it’s being included in a census of sustainable sources.


Turing is a Godfather of AI. But you can't say there never will be another. In light of recent achievements there should be new pioneers and Godfathers. Its the nature of innovation.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: