bustling-noose

Sounds like my next read. Thinking fast and slow was something I just started but it seems like something I want to read when I’m functioning better.

I agree. While depression and negativity leads you to see things practically rather than in an optimistic and hopeful way, the quality of life and satisfaction in a little optimistic world is much better. I am yet to read Daniel Kahneman but he talks about optimism a lot.

I would like to add to this. I have been depressed since I was a teenager. Anxiety Panic attacks, poor sleep, s*ci*l tendencies. I was able to finally get a job at 29 and first 3 months into it, I realized this is not a right fit and the company and its management is chaotic at best. It was a red flag right from the start but I ignored it because I was desperate for a WFH job. After 3 years of therapy, my views had changed that it's not so bad (something that I think the optimistic view changed). I was also looking to move to North America so I kept stalling to find a new job but that was a different thing. After I slipped back into depression a while ago, I again started seriously considering quitting because now in 3 years the company has grown somehow and some of the employees are really toxic. In 3-5 months I seriously want to quit this time whether I switch or no as I will complete 4 years at a company I never planned on working beyond 6 months. So there might be some truth to this. When I am depressed I see all the realistic things going on. When I am doing well I tend to ignore lot of the red flags.

I want to know if I can use an LLM during problem solving at an interview for SDE jobs at Amazon, Google and Microsoft who have boldly claimed lot of their work is done by LLMs. I will be able to point out the algorithm and explain how it works including the time complexity, but I will need the LLM first to solve it.

Will SDE interviews change ? Are these companies gearing up to let AI engineers in ? I highly doubt this is ever going to happen.

There is a quote by buffet that I think applies to a lot of scenarios not just investing : 'to be fearful when others are greedy and to be greedy only when others are fearful.'

ML IoT 5G blockchain etc. so many technologies are great that had their gaussian curve moment during greed. But these things take a back seat after that.

Definitely poor branding. It needs much better 2025 branding and it’ll probably take off in this AI search era where results are really poor tbh and very inaccurate at times.

Protectli makes great hardware. But unfortunately intel runs plenty of code with things like management engine that requires Coreboot to disable during boot. It may also be possible that the cpu refuses to boot with ME disabled so maybe coreboot doesn’t always mean doesn’t run proprietary code. True opensource will maybe happen with RISC-V when it comes to routers fast enough to be installed at home or small offices.

That being said, home routers are the least supported devices when it comes to security and privacy. People are running age old firmwares that are known to have exploits. These things are literally so cheap and poorly maintained anything with openwrt is going to be better.

For offices I would not shy away from recommending protectli with openwrt or opnsense as long as there are people with enough expertise to maintain these things long term.

I’ve been working with a team of so called ‘junior’ devs and coding is the least of the problems. Design and architecture is by far the most difficult thing for people to understand and get a hang of. A staff engineer won’t replace 2-4 junior devs with LLM. He will focus on the design and architecture and then get 2-4 engineers to execute it with LLMs. The two might sound the same but it’s not. And that’s the difference between understanding software engineering and coding. LLMs make coding easier, not software engineering.

In India Amazon makes little sense. Neither does kindle nor does local book store.

The reason for this is that both paper and printing here is cheap along with labor. The original author also licenses for cheap. The publishing houses however take a large cut increasing the cost of the book.

People get a hold of the epub and print them and sell them for 1/4th of the price sometimes 1/10th for new and even less for used.

The only way I see around this is digital libraries. Let people rent unlimited books (but like Netflix limited at a time) and take a monthly cut.

Music on iPod or a Walkman or a radio is great.

But have you ever attended live music shows ? Have you ever ‘felt’ the music ? Even someone at a local bar singing feels and hits different.

AI can never bring feelings. That will never change. Even science fiction agrees with that.

So bring all the AI you want everywhere, some things are irreplaceable by electronic world.

With the kind of funds raised by AI companies there should be 10 different ones sitting next to me competing with each other.

My theory : Engineers rented servers and maintained them with software and packages and infrastructure scripts / code etc. Then this moved to cloud VMs cause it became easier for higher availability and sometimes cheaper too. Then VM costs started rising and cloud providers started offering tempting prices to use some of their services instead that integrated well with their VM infrastructure. Simultaneously engineers started to cost more money to maintain these systems and more people trained in 'cloud' became available for relatively cheap. So people moved their infrastructure to cloud offerings. Now both engineers and cloud services cost a lot of money. But now engineers who can maintain such infrastructure are far and few in between and also cost a lot of money while cloud offerings became turnkey literally.

So costs went from $80k an engineer a year maybe a decade ago with a few thousand to servers to $200k an engineer which you would struggle to find or $100k for a 'cloud engineer / architect' plus $100k to a cloud provider.

This sounds great in theory. Except that cloud providers are messy and once vendor locked in, you are in a big spiral. Secondly the costs can be hidden and climb exponentially if you don't know exactly what you are doing. You might also get into weird bugs that could be solved by a patch over a Monday to some package you could just update which might take months or never happen on a cloud provider. The reality of moving to cloud is not as rosy as it sounds.

Universities used to be the birth place of big projects that were created to solve problems they ran into hosting / running their own infrastructure. I hope that is still true.

While the majority of revenue of Microsoft is not Windows anymore, I think Windows defines the brand much like how iPhone and Mac define Apple even though that might be part of the revenue not all of it.

What I am curious about is what happens when the original product that makes the company popular starts to experience poor quality. Take Google for example, its search has been on a decline in the last decade or so and needless to say the company is experiencing problems as well in the last few years. While GCP and GSuite are significant, people have lost faith in Google which probably started with search.

Windows 11 and the iPhone seem to be heading towards same fate as Google search imo.

Now if only we could see inside a black hole.

Very simple:

Big companies like Swiggy and Zepto will mine the F out of your data. Some of it is for their benefit but some of it they could sell in the future. These so called founders are really just another wolf of app street looking to pump and dump. So when they do dump, or when some VC comes with money, they don’t just sell their app they sell it as a whole package of data and analytics that some company can use to sell their product or something VC can leverage to sell their stock to someone else. It’s not that difficult.

As far as smaller apps go these apps outsource their development to people who come with ‘packages’ to develop and maintain their app. These packages are the same logic as above but it’s just that they come from some template so you might be asked for location permission or camera or microphone by some really random app that has nothing to do with it.

While the quality of iOS is degrading, some of these things are really important and simply work better on iOS.

I was about to type: the man could have done so much for this world. Instead he chose the dark side. What a sad story.

The quality of dns always makes or breaks your internet experience. Personally at home unbound on opnsense with some blocking list has always worked really well for me. Openwrt with pihole also works fine. But the moment I have to use some recursive dns like this, I tend to not enjoy the experience.

Correction here: it wasn’t apples AirPods, it was bose quit comfort over the ear iirc. That’s why I could hear the other person. But I think they could hear me cause of maybe both meet and good mic array.

I was sitting at a Starbucks next to a VERY noisy street on a google meet call on an M1 Air with usb-c AirPods (the cheap 19$ one) and I asked the person on the other end if they can hear me at all. To my surprise they couldn’t hear any noise just my voice. No idea which part in the whole setup achieved this but I feel like stuff like AI and all have some applications that can blow you away. Not putting the damn thing in everything!

Yes it seems like this is something to do with hardware testing. Maybe memory or registers or something that needed just X bytes etc for overflows or something. It’s really random and the only person who would know it is the one who wrote it :)

And many companies like to have a portfolio of both, or an offer at least from the closest competitor. Tim at Apple for example played this game for decades to make sure he always received competitive prices for parts. They tried to pull it off with modems and it backfired badly. So they built their own. But if QC was to play really nice with companies like Apple and Samsung etc, they could actually maintain their moat. But then good management is another important requirement which is a problem here.

>though nowadays they have more competition in that area

This is what is called lack of moat. If today, every router beyond 1Gig was made by Cisco then the moat would be so wide that the internet would literally not exist without it.

The three things that any value investor values in stocks - Great management, good reinvestment of capital and a really wide moat. Cisco didnt have a wide moat. Nvidia currently does, but it may go away sooner than later as they don't really have any special sauce to their graphics cards as such. Same goes with OpenAI, or Qualcomm (Apple made a modem now) or any other tech company with an illusion of a wide moat. Tech businesses only have a wide moat until someone decides to compete seriously and that can usually be done with a lot of money in a relatively short time.

Some of the businesses like Walmart and Cosco for example have a really wide moat in the sense that their business is not something that someone with just money can beat in a few years. Look how amazon setup physical stores with money and failed. It's extremely difficult to setup that kind of business. Trust from suppliers and consumers takes decades in those kinds of businesses.

If your business is relying on a moat that someone can beat with R&D and money in a few years, then you don't really have a moat.

What Tim Berners Lee doesn’t know or understand -

Social media is like fentanyl or cocaine for the masses. It’s preying on people’s stresses and poor mental health by providing temporary relief or disconnecting from the real world. Social media being a place to connect and maybe even run a small business is like 1% of it. It’s a place for attention seeking people filling their loneliness with fake connections and content while content consumers are trying to achieve the same buy consuming instead.

Ive seen people unable to get out of instagram. Their fingers are constantly twitching and scrolling. This is just like drugs and controlled substances. The government punishes the dealers and is trying to control but failing cause there are plenty of corrupt people willing to bypass this. Same goes with social media. Zuckerberg is probably worse than Escobar.

Social medias promise of social connections was really just the same as what drug dealers offer, the good stuff for new consumers and then once they are hooked give them the cheap stuff. They really don’t care about ‘social’ part of it because the addiction potential is so high.

Start doing mass studies on cognitive problems created by instagram and TikTok. Publish them in big journals and show how social media can go from social to addictions so quickly. People already consider phones as part of their body, I won’t be surprised if people also consider their instagram account more valuable than their health.

I think by fuse I mean't its stuck on to the SOC module, not part of the SOC as I may have worded. While you could maybe still add NANDs later in the manufacturing process, it's probably not easy, especially if you need more NANDs and a larger module which might cause more design problems. The NAND is closer cause the controller is in the SOC. So the memory controller probably would also change with higher memory sizes which would mean this cannot be a last minute change.

Sam Altman should have sold OpenAI to Musk for 90$ billion or whatever he was willing to pay (assuming he was serious like he bought twitter). While I find LLMs interesting and feel many places could use those, I also think this is like hitting everything with a hammer and see where the nail was. People used OpenAI as a hammer until it was popular and now everyone would like to go their way. For 90$ billion he could find the next hammer or not care. But when the value of this hammer drops (not if it's a when) he will be lucky if he can get double digits for it. Maybe someone will buy them just for the customer base but these models can become obsolete quickly and that leaves OpenAI with absolutely nothing else as a company. Even the talent would leave (a lot of it has). Musk and Altman share the same ego, but if I was Altman, I would cash out when the market is riding on a high.

The MacBook Air with M4 chip 16gb ram and an amazing display and camera is just 999$. During back to school that will happen soon it’s 899$ with free AirPods. That’s really great value given how good the hardware is.

I wonder if Apple needs to reconsider Xserve. While Apple probably have some kind of server infrastructure teams, making their own server infrastructure out of their own hardware and software sounds like something they could explore. The app ecosystem coupled with apples servers offered in the cloud or ones you could buy would be a very interesting service business they could get into. Apples App Store needs better apps given how much the hardware is capable of now especially with iPads using M chips. A cloud backed hardware and software service specially designed for the app ecosystem sounds very tempting.

The hardware has evolved faster than software at Apple. It’s usually the opposite with most tech companies where hardware is unable to keep up with software.

My thoughts too. This product was in the pipeline maybe 2-3 years ago. Maybe with LLMs getting popular a year ago they tried to fit more memory but it’s almost impossible to do that that close to a launch. Especially when memory is fused not just a module you can swap.

>We are deprecating the blocking version of the webRequest API. This required extensions to proxy all network traffic to provide filtering capabilities, which came at a performance and privacy cost. The new declarativeNetRequest API provides a safer alternative for many use cases.

This is from manifest v3 google page. Is this declarativeNetRequest API not able to provide any filtration ? Proxying traffic does affect privacy, I agree, but that also means that Google is trusting all traffic by default which is another privacy concern. So the privacy concern seems to not make sense except that in one of the two, google loses money because of ads being blocked.

The article almost got it right ‘emotions, peer pressure, temptations’ and how people spend money on things they don’t need and then got it wrong by not actually saying what’s important.

Money management and financial literacy involves psychology along with math. So does sex education like the article compares it to. Having sex or spending money is not the problem. Nor is putting fear of the two in your head from a young age or the desire for it.

As Charlie munger said - psychology should be taught from a young age. Maybe the most important topic missing from education.

Spending money isn’t bad and having sex isn’t either. But saving for a future you’ve dreamt of and sex with a loved one is better. This involves less math and sex education and more psychology and mental well being.

If you are a 30 something politician looking for a change and understand psychology, you might want to visit schools and give children hope and mental well being tips. They might vote for you in 15 years.

Providing encrypted data and not providing encryption are two different things. The CLOUD act requires you to hand over data. It could be encrypted. The UK government is asking to hand over data that is also not encrypted. The two are not the same. Note : Not American.