paxys

The "without any direction" part isn't correct. Sure they may not have been explicitly told to do it in this specific prompt, but dig through pre-training, post-training, reinforcement, alignment material, fine-tuning, system prompts, tool calls and more and there's definitely very specific training and instruction for how to behave.

Well yeah the copyright violation is pretty clear cut here.

Thinking of an algorithm in the abstract is one thing, implementing it at scale is another.

Yes you could have invented PageRank, but could you also have invented MapReduce, BigFiles/Google File System (GFS), Google Web Server, Bigtable, Protobuf? Then spun up fault-tolerant clusters consisting of cheap commodity PC hardware in an era where AWS wasn't even an idea yet? Then invented the concept of Borg to manage this hardware globally?

It’s ridiculous that so many issues affecting this country over the last two decades can be traced to the inability of Congress to pass laws.

For context Meta had $61.8 billion revenue and $15.85 billion net income in Q2 2026. So this entire “existential threat” of teenage safety amounts to one bad quarter for the company, and they will continue on unscathed.

Almond farming - ~1.5 trillion gal (1.1 gal per almond)

Or more simply – $25K (+ tax) put in a savings account will earn about enough interest to pay for a $100/month AI subscription indefinitely. And at the end of it you still have the $25K.

Probably worth it for Ollama but you can run Jellyfin off a raspberry pi.

More validated by the day that my $450 M4 Mac Mini (16GB) was the best deal in computing for a long, long time.

While they’re probably right from a consumer perspective, the article skips over the fact that accepting and handling cash is a significant cost for businesses, way more than the credit card fees. Delays in checkout, making change, counterfeit bills, employee theft, external theft, safe transport, added accounting burden…all add up to an estimated 5-15% (https://plainscapital.com/blog/the-cost-of-accepting-cash/). In fact merchants can now legally pass through credit card surcharges to customers but very few choose to do so, because they’d prefer you pay by card.

In that case a credit card is actually significantly better than debit

Delivery apps don’t mandate that the price on their apps be the same as on the actual menu. If you walk in and order you’ll pretty much always get a lower price.

One isn’t more real than the other. They are both numbers in an online database. In one case your assets are going down, in the other your liabilities are going up. The net result is the same.

Can we audit American corporations?

No, the whole point is the e-ink screen. Kindle has had a web browser since the very first version, and people have managed to read books on it just fine without distractions.

It’s a pretty simple explanation.

Bay Area does not have much industry outside of tech. New York City is a hub for banking, investing, insurance, government, healthcare, consulting, real estate, fashion, law and more. All of these are now increasingly hiring tech workers directly and also retraining and transitioning their existing workforce to be more technical.

So just by numbers alone Bay Area will never maintain a monopoly on tech jobs, especially considering the NYC employment market is several times the size. However if you just filter to the tech sector the numbers will be very different.

Computing is a task. Printing is a task. It’s not wrong to call both a human and a machine a “computer”, because computing (applying an algorithm to an input and producing an output) is literally what they are doing. Same way you can have human and machine diggers, cleaners, calculators and lots more.

Anthropomorphizing comes in when we attribute much more complex behaviors to them - chain of thought, reasoning, intelligence. At that point you aren't talking about mechanical concepts but claiming that these machines are exhibiting human behavior.

The person paying you doesn’t care about your enjoyment.

Before LLMs we had a pretty good idea of security boundaries in software. Applications didn’t trust user input. Operating systems didn’t trust applications. Services and processes didn’t trust each other. There were always tokens, scopes, delegated grants.

Suddenly every AI company’s security model seems to be to say “pretty please” to a non-deterministic machine and hope for the best. And if there is a security failure instead of accepting blame they go “well we can’t help it, our model is too intelligent”.

I feel at this point all new AI announcements can be auto generated by AI.

Next up - Slack releases a model router. Cloudflare builds an agents platform. OpenAI acquires a dev tools startup.

I used to be that person, and suddenly I haven’t opened an editor or written a line of code by hand in 6+ months.

“Paste this AI generated text to your coworkers telling them not to paste AI generated text to you.”

Not all slop content is AI generated. Remember the glory days of Buzzfeed? Stock images surrounded by puff pieces is quality pre-AI space filler.

AOSP is open source. The Android running on most manufacturer-branded phones isn’t.

Doesn’t Apple do the same?

Google uses libraries licensed under GPLv2 in Android (I’m not sure which specific part of Android the author is talking about), and so is required to make the full source code available for anyone to view. They previously used to publish release tag tarballs, but now require you to fill out a Google form and then (weeks later) will share the source with you on Google Drive.

And what do you do when you find them? Unless there are adequate consequences (and no, a paid vacation or getting “fired” and moving to the next town over don’t count) the abuses aren’t going to stop.

A simple website that conveys all the information in this one would have taken less than an hour to make. There’s no need for all the bells and whistles for a project like this.

> and indeed the acquisition price of a company is almost never exactly its “market valuation”

The acquisition price is almost always significantly more than its market value. The reason is that if you start buying shares on the open market each transaction of yours will raise the stock price, and eventually you will price yourself out of the purchase.

So what buyers do is negotiate directly with the board and propose an inflated price that >50% of shareholders are okay with.

Items don’t have an inherent “worth”.

If I have an apple and you agree to buy it for $1T, then the apple is indeed worth $1T at that moment.

After the purchase is complete it is worth ~$0, or whatever the next person is willing to pay you for it.

Alternatively you can say “it is still worth $1T to me” and that would be true as well.

When we are specifically talking about companies, their value is where buyers meet sellers in the order book.