citrin_ru

If LLMs is just another technology then low end will eventually win. The bet which I guess many AI investors are making is that LLM will allow recursive self improvement which will lead to development of super-human AGI. The company which will get there first will rule the world thanks to an immense power enabled by AGI. There will be no 2nd or 3rd AI company, the 1st one will wipe the rest out.

The example shows an auto-incrementing user ID. How’s that possible without contention between all shards?

Why not divide ID range (63bit?) by the maximum planned number of shards and then set on each shard it's min/max value so ranges will not overlap?

Old generation DIMMs are not that much cheaper and supply is limited as old generation are taken out of production to free capacity for newer chips. DDR2 is already more expensive than DDR3, likely because it is no longer produced but there is still demand (to replace/upgrade memory in older hardware).

Outsourcing never ended but it did happen in waves. Right after ZIRR ended but before the AI panic multiple companies started to layoff in developers in western countries and double down on outsourcing to India. During ZIRP they employed more people in the US/EU and less work was outsourced.

May be LLM resume screening is a symptom of a bigger problem - with tens of candidates per vacancy employers can screen resume badly and even throw half of the resumes away and still hire someone qualified.

Threads are not free - they allow to use more CPU cores but if you load is not too high than with a single thread memcached uses less CPU than with multiple.

What we see in quotes is not the price of oil being shipped right now; it is usually a one-month futures contract. Traders expect that by this time, Hormuz will reopen. If it does not, they will lose money, but the actual price could be much higher.

If memory shortages were not caused by AI then may be (but still unlikely). In the current situation even cautious in the past companies are switching to “move fast and break things” mode to not fall behind vibe coding competitors. When the velocity is the top priority there is no time fo efficiency.

Web browsing is the most resource intensive task given how bloated modern websites are. It’s easer to imagine a cheap phone with a set of optimised native apps but the problem is tech giants are not interested in reducing footprint of their apps at the same time actively hostile to 3rd party clients.

Surprisingly high number given that people are being told by tech CEO that AI will replace all white collar jobs soon and a few years later AI guided robots will replace all blue collar jobs too.

Many managers are more bullish on AI and less able to recognize slop, they are unlikely to recognize quality crisis. And they are the people who decide who belong to the industry and who is not. As a result we will get an escalation of enshittification and people will start to forget that slop is not the only option.

AI investors bet on the high return in the next 10-15 years, they have no reasons to care if everything which require semiconductors will become much more expensive in the process.

A typical modern car is already hideously complicated and a different type of motor would not change this.

What is the current market sentiment? Share of EVs is slowly rising so having a good motor as important as ever.

Gas prices vary across the world, mostly depending on if it's only LNG imports (high prices) or local sources (much cheaper, e. g. the US) or a mix (e. g. Europe but since 2022 it's heavy reliant on LNG which pushes prices up). Coal prices also vary. Guess we cannot compare gas and coal prices globally, only withing a specific region.

A big threat for illustrators is not LLMs per se but slop acceptance. We've seen the same with machine translation - it was (and to some extent still is) worse than human translation but people accepted it because it's free. I recently seen and ad on a street which caught my attention - it was a photorealistic image of a taking off plane. It looked off and when I looked at details it's become clear it's AI generated - the 1st giveaway - stationary turbines on a flying plane, an angle relative to the runway is wrong, small details, especially in the background are off. Someone prompted LLM, seen the result and decided it's OK to print this advert. IMHO they would be better off with buying a stock image.

In regions where winters are dark windy places tent to be not too far away. If both solar and wind both overbuild to the extent it makes sense (say 5x for solar and 2x for wind) and batteries cover a normal daily cycle you probably will need to burn gas on 10-20% of days which is not net zero but way way better than the current situation.

Gas allows to use combined cycle gas turbines (CCGT) which is more efficient and it makes gas cheaper for electricity generation.

It's a good news but I didn't expect that coal is still on the 1st place and not really trending down. I though coal was largely replaced by gas years ago...

I'm noticing some decline of skills I don't practice regularly and LLM is just one of reasons why one stops practicing. Switching to another area of work gives a comparable decline. If you want sharp skills you have to use them.

IMHO peak GUI was in 2000s - on Windows most app used Win32 API and apps which followed "Microsoft Windows User Experience" guide had consistent UI/UX. Since then Microsoft introduced many competing frameworks to create GUI all look slightly different and UX is less consistent too. And then Electron come which brought inconsistency of web to the desktop apps.

TUI returns what GUI lost over the last decades - low latency, low RAM/CPU usage, ability to navigate without a mouse. No one seems to be building nice GUI apps anymore.

Nobody ever got fired for buying I̵B̵M̵ AWS. Most corporations already use AWS, used to its legal terms and accepted the risk. Any new provider will be scrutinised by legal more than an existing one.

I'm personally was fine with contributing to open-source without any financial reward. But I'm reluctant to release anything in public now because it will be eventually incorporated into the training set for the technology which will (or at least can) lave me without a job and chances to find one.

A significant job loss will trigger a deep recession which will eventually hit most AI customers so they will have too few customers to be profitable. The best (for AI business) scenario is when productivity is increased without mass unemployment.

The problem is that many companies which had reasonable leadership in the past with the advent of LLM AI started to make rushed (and dubious from my point of view) decisions - using token usage to evaluate an employee performance is just one of them.

It's hard to separate impact of workers being replaced by AI from an impact of a recession (or stagnation). It's not obvious how AI impacts employees with various experience: on one hand a senior is better at spotting AI hallucination one other hand a junior using AI can do much more than a junior was able to do a couple years ago for a lower (if adjusted for inflation) than a couple years ago salary.

IMHO it's around 4 years: inflation surge in late 2021/22 and following years was not matched by salaries but 2020 was not yet that bad.

Why we should assume that whoever offers these Chinese models makes sufficient profits and will not rise the prices eventually too?

At least partially the usage is driven by free plans. I use Gemini and ChatGPT for free. I will not pay for them unless traditional web search will be killed (google quality is subjectively on a downward trajectory for the last few years). My employer pays for AI but IMHO it's driven by a panic level FOMO, not evidence.

It seems the more people spend online the more they trust online picture whatever it is (for most people it's an algorithmically generated feed) and less pay attention to what they can see in person. Smartphone addition is harmful in so many ways.