Search for one is absolutely horrendous. It used to be great, but not so in the last years. Nowadays it’s filled with spam sites that they don’t seem to be able to filter out. And don’t get me started on the crappy AI overview that hijacks all queries.
Just today I tried a query for water filters and 1/3 of the results were ads. The other third were product pictures, or businesses in close proximity based on my ip. Then there was a box with related products/services, which was completely irrelevant to my needs, a second box with places, yet more product images and so on and so forth. Practically 70% of the real estate of the page was occupied by things I didn’t ask for. All I want is a list of relevant sites to go there and judge for myself. I don’t want Google to spoon feed me.
Raspberries are basically for DIY projects. I have one on my router handling call blocking for my landline. If it was costing $300 I would rather get a mini PC or use one of my defunct phones with UserLand on it. I can't see any world where a comparison to an entry level laptop makes sense.
Anthropic or OpenAI have no foothold into the mobile market. Google has integrated a shitton of AI functionality into their latest Pixel phones. That’s what would scare me if I was Apple and worried that if AI prevails this could steal some market share from me. The other two are irrelevant in this context.
I dont think so. These entities and the hardware they own would be bought for legitimate AI use long before they'd hit the open market. AI is very useful, and even profitable at the inference level. It's just an open question whether this monumental amount of spend for research is worth it.
The DDR5 will be registered DIMMs. The GPUs will be 600W paperweights with a custom form factor. Similarly the NICs and other PCI-E accelerators. The motherboards also adopt custom form factors to fit in racks. The hard drives will be using SAS connectors. The flash will be in E1.S form factors.
The server CPUs that you want for a home desktop or small server, high clock SKUs, will be in high demand.
Any savings for someone willing to build a system from second-hand server hardware will be eaten by using adapters or sourcing a rack.
I'm not saying you won't be able to make a slightly outdated frankenserver with more compute than you need, I'm saying that's not going to bring down prices for Grandma's machine that she needs working to check on her retirement account.
Even if we assume that everything you said holds true, how is that we as a crowd can make viable a service that eats some $300bn annually in infrastructure costs? Where would that money come from? Most tech companies these days are cutting their AI budgets because the per token pricing is killing them.
Cite a real source for that last bit, I don’t think that is true. Also the budgets should be cut the spend at some places goes beyond any reasonable amount. The strategy there is to hook everything in and find the right processes, then cut the rest. Things then get better and better with each model release.
The way you make a viable service that eats 300bn annually is to have enough demand to service that. Anthropic underbought compute. That tells you something.
I think over the years local models have fairly consistently been ~7 months behind frontier performance. Local models are hugely important but I don’t see the calculus changing. I can imagine it’s certainly the case for many tasks that there will be diminishing gains for performance improvements or reliability pass some threshold, in which case you don’t need frontier performance and you can certainly use local models or at least cheaper tiers of proprietary models if local is too much of a hassle. Plus of course use cases where local is necessary or the pros of having local models or on device models outweighs that of frontier.
Maybe things will change though, I would assume through basically government subsidies from China etc, to undercut existing frontier labs, but you can always spend more (better data more compute etc) for better performance and that I can imagine will always have a selling point.
Nothing, they just work as intended. I bought a Fujitsu A/C that supposedly required registering through some app. Never connected it, and I removed the wifi kit at installation. Works like a charm no problem whatsoever.
I can only speak for myself. I mostly write C# and the code they generate isn’t of my liking. Too many var statements, async everywhere even for the most trivial things, and then a lot of times when I ask something difficult they fail spectacularly. I asked Claude the other day to propose an algorithm for Named Entity Recognition for brand names and all it did was to build a heuristics based on the type of business entities. And then in order to do that it went and created four classes. They overcomplicate things all the time for no apparent reason.
I built an Asterisk dialplan using Gemini a few months ago and it has so many hacks and quirks in it that I don’t dare touch it. I just leave it there hopping that it will keep working. I excluded Asterisk from future OS updates to avoid something fucking up that travesty of a dialplan.
Couple of weeks ago I wanted to explore an enhanced version of a PPMI algorithm that tries to flatten rare pairs. I went back and forth with all three major LLMs (ChatGPT, Claude, Gemini) and they couldn’t agree on anything. They kept changing each other’s algorithms all the time. The “reasoning” part seems laughable, it’s just a probabilistic infinite loop.
And that’s just for the programming part. Whenever I’ve asked them about SEO their answers are highly inaccurate. All in all I just don’t fucking trust anything they say.
There’s nothing baffling to it. Windows PCs are upgradable. Apple won’t even give you a PCie slot on its $10k mac studio ultra to install a better network card or whatever.
I haven't worked with TOO many different companies, but I have worked at a few of various sizes (from small startup to huge Fortune 100), and none of them ever provided upgrades for machines. It was always full replacements. Sometimes you would get a used machine, but they were from someone else who left, not an upgraded machine.
Are other IT shops really doing a lot of piece by piece upgrades for employee machines?
As a recent retiree from a Fortune 500 company...no, there's no such thing as an upgrade. We were virtually exclusively laptops on the desktop. It was full replacement every time.
> Are other IT shops really doing a lot of piece by piece upgrades for employee machines?
I doubt it. I'm certainly not, and none of my peers at other companies locally are either. Even less so now that plenty of business class laptops are coming with soldered ram anyway. The MO is to just replace the machine once its out of warranty.
If your standard developer laptop has a 256 GB SSD, but a certain team needs more disk space due to the work they're doing, you can just add a second nvme for a fraction of the cost and inconvenience of replacing the whole laptop.
> "Are other IT shops really doing a lot of piece by piece upgrades for employee machines?"
My experiences cover only Europe, mostly in sasec (safety and security, not infosec) shops, including sasec-related engineering and product development. The only Macs I see in any pro capacity are those of clients and rent-a-lecturer/instructor-types, the latter seldomly part of the industry. In my neck of the woods we run mostly on machines from Panasonic and Lenovo; in-house repair labs are a thing (some of them with expertise and equipment that makes the Rechenzentren at the local universities bow their heads in shame).
What a lot Apple people don't seem to get into their heads is that there's user segments to whom the virtues of Apple's "silicon" is utterly irrelevant; the small benefits you'd get out of it are completely negated by a litany of cons that makes their products completely undesirable.
There’s nothing baffling to it. Windows PCs are upgradable.
We're talking about enterprises here, not home tinkerers.
Enterprises buy whole computers and replace them every x years. They don't waste expensive IT employee time running around upgrading machines all the time.
The last time I worked for a company that did any repair of its computers was around 2005, when all ~500 Dells in the office had to have their defective motherboards replaced.
What company upgrades their windows PCs? They give them exactly as shipped. IT department is not wasting time swapping out RAM or SSDs. And they certainly are not upgrading them over time. You just replace the entire PC if you go to 'swap' it.
I work for a little company called Boeing and all our PCs (desktops and laptops) are Dell and our IT center will upgrade SSD, memory, and even do repairs like swapping out motherboards.
Probably helps the IT center folks are actually employees of Dell and this service is part of the deal Boeing has with Dell. Lots of big companies have similar deals with their hardware vendors.
Interesting. I’ve worked for company’s that were all Dell shops as well and were similarly large and they had no such deal. You got whatever PC was available (like two options, one large and one small) and no choice beyond that. If you were special and could adequately plead your case, you maybe got a desktop for the extra RAM which was special order.
Why would you even want a SSD or memory upgrade? By the time you’re out of memory, the cpu upgrade is typically worth it.
A month or so back I was hanging out at the IT center waiting for my laptop to be re-imaged (my Windows bootloader went "missing or invalid") when an engineer brought his laptop in to be upgraded from 16GB to 32GB. The IT tech took it in the back, did the job, and brought it back out in about 10 minutes.
I think part of the problem is all our laptops have smart card slots on them, and that limits the available options.
> Why would you even want a SSD or memory upgrade? By the time you’re out of memory, the cpu upgrade is typically worth it.
Not necessarily. Sometimes the default laptop sizing comes with a standard usage in mind but more space and ram is justified for other roles. Sure you could have different laptop models but if you are fine with just more ram and disk space why not?
I have had an upgrade maxing out the memory of my works thinkpad. It was a number of years ago, 2020 or 2021. Might be less common these days for obvious technical reasons (solderd ram on many models) but if the hardware allows it why not?
I work I Ag retail (agronomic services, chemical sales) and while we have an IT department, a decade ago I’d occasionally act as “local” tech support and double a coworker’s RAM when their combination of browser/office/database front ends stopped gracefully fitting in 4GB (and later 8GB). I would also migrate them from HDD’s to SSD’s, and set them up with backups.
But even I haven’t done that in several years now, once IT moved to providing 16GB memory and SSD’s as a baseline, there’s really nothing left in a box to upgrade. I’m quite happy enough to not have to care.
Something as simple as adding a stick of RAM might be worthwhile, but some upgrades will take more money in salary (you and the person whose computer you're working on) than the upgrade is actually worth. This is especially true if you replace several components one at a time in a what would otherwise be a single replace cycle.
This is especially true if the business is writing down the replaced hardware as depreciated capital, compared to say simply adding a stick of RAM.
I've generally been given additional decent-performance storage every year or two when working in video games, fairly reliably the case since 2000, seemingly regardless of company size. Somehow each new project would always require yet more working space (again), and projects would often overlap to some extent.
Aside from that, almost always a new PC - almost! When times haven't been so good, I've had spot upgrades.
All of the above is just a question of running the numbers and minimising for cost, I'm sure.
On very rare occasions I may do that for a user, if I happen to have extra RAM on hand from - for example - a broken machine. But by and large it's just going to be a whole machine upgrade.
Just today I tried a query for water filters and 1/3 of the results were ads. The other third were product pictures, or businesses in close proximity based on my ip. Then there was a box with related products/services, which was completely irrelevant to my needs, a second box with places, yet more product images and so on and so forth. Practically 70% of the real estate of the page was occupied by things I didn’t ask for. All I want is a list of relevant sites to go there and judge for myself. I don’t want Google to spoon feed me.
reply