partial_accumen

5KB of RAM

TIL! I’ve been living a lie for decades thinking the 20 in VIC-20 was the amount of RAM like the 64 in C64 meant the amount of RAM. I only owned the C64, never a VIC-20.

Commodore VIC-20 came out in 1980 with 20KB as well.

This article is insane and proves me more that AI is product for rich people. Most of developers won’t see $100k per year paycheck in their lifetime.

These days an annual salary of $100k is at the very low end for most IT jobs in the USA (beyond the junior level). Even in my MCOL area $125k-$200k is more common.

You burn 16 tokens and what do you get?

Another day older and deeper in tech debt.

If power generation becomes so cheap that it can’t sustain the company then don’t rely on that for revenue.

I’m not aware of anywhere power generation is that cheap yet. That may be a problem for the future when commercial fusion is viable, but thats likely a lifetime away.

I’d rather pay a flat rate for the infrastructure and operating costs than a fluctuating generation charge.

I think everyone would, but the cost for generation is always fluctuating because the variation in the market for the fuels that generate electricity, supply, and demand of electricity on the market. If its a flat rate, and that rate is below the cost of generating the electricity, who pays?

I mean good-ish in the lesser-evil type of thing. I don’t expect any of those to be 100% ethical but there are some that are a lot worse than others

Ethics are subjective. “Good-ish” to you may mean you’re fine if its trained on copyrighted works as long as it wasn’t done with electricity from diesel generators belching exhaust into the local Memphis atmosphere (I’m looking at you Grok). Llama doesn’t do the diesel generator thing, but its a product of Facebook corporation. So is that “good-ish” to you or not? I don’t know. That’s up to you.

It may not be fast, but your i3 laptop with 12GB of system RAM can absolutely run a local LLM. This is where that “performance/accuracy” question I raised comes in. It won’t be very fast, and you won’t be able to run the most common large models like GPT-5 etc. However, if your needs are light, light models exist. Give this a read

Depends on your definition of “good-ish”. Do you mean:

• performance/accuracy?
• ethical origin?
• ethical ongoing operation?
• privacy/future data harvesting concerns?

Running one locally on your own hardware would likely reach “good-ish” with some sacrifices against performance/accuracy (unless you’ve got a lot of expensive hardware to run very large models). As far as ethical origins, there are few small models trained on public domain/nonstolen content, but their functions are far more limited.

Woah there buddy, the start of the line for the 5 1/4" floppy is back there. No cutting.

4116s are DIPs. I’d de desolder those myself for installation into my Intel 8088 luggable.

What I’m surprised hasn’t happened yet is RAM ICs being recycled at the retail level. As in, you could bring in an old laptop or phone with 32GB of soldered RAM and it would be desoldered and sold for cash or possibly even soldered into a new device you buy from that retailer.

I wonder how close we are to that business model arriving.

Under which rock does he live?

Under the rock where reading comprehension exists apparently.

Where he was prompting for “What is 2+2?” to the LLMs, the accuracy of the answer was immaterial. At that step he was comparing two systems and simply needed a static question to give both system to compare the internal processes to determine why they arrived at different outputs (or a what appeared to be race condition/infinite loop for one) when the result should be identical to both irrespective of how right or wrong the answer is to the prompt. The LLM answer from the LLM could have been “ham sandwich” and it still would have served his purposes.

What is the alternative to existing giant companies investing in the infrastructure? And why hasn’t that alternative already addressed this issue before the tech companies arrived?

I like the essay’s highlighting European contributions to software and technology, but it doesn’t quite answer the fundamental question of its title:

Why there’s no European Google?

The essay’s answer is [paraphrased] “…because we don’t need it.”.

I don’t quite understand that position because if a Google wasn’t needed in Europe then Google could disappear from Europe and no one would notice or care. Yet that isn’t likely the case. If Google disappeared overnight it would likely have massive impacts on business and personal lives across Europe.

I guess my answer to the article’s question as to “Why is there no European Google” is that creating Google (or a European Google) is extremely resource and financially expensive. Unless the funding for that effort comes from somewhere, it won’t just happen in Europe spontaneously without replicating the same private business model that many dislike about Google.

P.S. Another European created technology that should be added to the list for accolades is the creation of Deepmind machine learning/AI. This also lead to the creation of Google Gemini. While this is owned by Google, it was created out of the London offices.

Space: Ivashov describes how the last working manned launchpad at Baikonur was accidentally destroyed due to poor maintenance. Russia in effect can no longer deploy men into space.

Its even worse for Russia on this point. Not only was that the only crewed launch pad to the ISS, its also the only one that Russia has to launch uncrewed resupply missions (called Progress cargo spacecraft) to the ISS. So the only cargo ships going to the ISS now are from the USA and Japan.

On the downside, Energy Dome’s facility takes up about twice as much land as a comparable capacity lithium-ion battery would. And the domes themselves, which are about the height of a sports stadium at their apex, and longer, might stand out on a landscape and draw some NIMBY pushback.

This is surprisingly good! I would have figured it would have taken far more than twice the land than a Lithium battery solution.

I was thinking about much larger scale bubbles in “unwanted” geological depressions such as old open pit mines or rock quarries. The depression in the ground might offer more protection allowing it to scale up higher in volume.

In every version of Windows up until now which has contained a taskbar and start menu, as far back as Windows 95. Not just Windows 10.

Sadly not true. Microsoft removed the Start button in a version of Windows before. It was in Windows 8 (and Windows Server 2012 for some godforsaken reason) with the cursed “metro” interface. MS did it for the same stupid reason they’re citing here “tablet and touchscreen users”. The uproar caused MS to release Windows 8.1 a year later where they returned the Start button.

but I think the realistic reading is it was simply a kickback to fortune 500 companies that got these politicians elected.

If there were no legitimate geopolitical reasons, then the “simply a kickback” would be much more plausible. Also, if it was a single source company, then “simply a kickback” would look true. Additionally, if was perhaps just domestic companies “simply a kickback” would certainly be even more likely. Lastly, the Chips act wasn’t just about production domestically. It also blocked sales/exports of completed high end chips and chip making equipment to China. If the Chips act was “simple a kickback” you wouldn’t do all that other stuff, and you certainly wouldn’t allow foreign winners (like Taiwan’s TSMC).

Was their rewards because of industry lobbying? Certainly. However, unless you’re in a purely communist system of government where all the companies are owned by the state, you’re always going to have private companies benefiting from government spending, tax breaks, and subsidies. As to this just applying to fortune 500 companies, there isn’t really a “mom and pop” semiconductor industry making handfuls of chips at a time except outside of engineering sample that are used in R&D for fortune 500 companies.

The worst of it hasn’t happened yet. The point where consumers can no longer afford to consume is coming.

Its mostly already arrived.

“As of June 30, the top 20% of earners accounted for more than 63% of all spending”

source

This means that the other 80% of Americans represent only 37% of the spending done today. If a company is looking to maximize profits the typical path is to do so by marketing to the group where they could earn the most money. That is less and less the bottom 80% of Americans.

The creator in that video seems to think the Chips Act subsidies were to benefit consumers by having affordable memory produced domestically. That wasn’t the goal. The goal was to derive drive GDP by having another source of domestic production, and drive job growth/tax revenue from workers working at the domestic facility. Lastly, it was to have strategic domestic production decoupled from other nations so we, as a nation, could not be held hostage by another nation (like we do to so many other nations) for crucial (pun very much intended) resources we need.

Nothing about that is about making RAM cheaper for retail consumers.