HN user
Ages ago, when secretaries (usually female) were still common in offices, my boss told me "learn the names of all of the secretaries and say hello <name> every time you see them. If they know you and like you, they will probably do you a favor, when you need it. They are the important people running the company."
You have to do visible, great work.
Apps that create false data to crash advertising algorithms. Random GPS locations, random product interests, random web site visits, random follows and likes. Like how SEO poisons search results.
Apps that respond to video ads by calling/messaging the sponsor CEO and the media company CEO with the audio of the ad.
Triplebyte developed an online assessment technology for programmers. The company was acquired by Karat. Their website has claims of enormous improvements to the hiring process. Previously, you could take their tests and use the score to add to your job application. https://connect.karat.com/tb-welcome
Advertising cost is added to the product price. In order to generate one trillion dollars in advertising revenue, there would have to be about 20 trillion more in product spend. Civilization does not have another 20 trillion sitting around to spend.
Currently, consumers are totally bombarded with ads, on TV, on streaming, on social media, on websites. There are multiple copies of the same ad during an hour time slot. Consumers go from slight interest, to irritation, to disgust, to nauseated.
Currently, the algorithms that target ads are complete failure. At least, they could get my gender right.
The next big advance in AI is going to be ad blockers for social media, streaming video, live television. Then the next step will be bots that respond to ads, with total click through rate and messages directly to the CEO of the company. Imagine an App that calls the CEO and replays the audio of their video ad. Or an App that DM's the CEO for every social media ad. The CEO bothered a million people, let the million respond directly to the CEO.
2001: a Space Odyssey. I saw it in the theater in 70mm when it was released in 1968. It totally blew me away. A majestic vision of the future. Thirty three years from now (1968) there was going to be all this marvelous stuff: Pan Am space craft going to a huge rotating space station to a round globe ship to a moon with several lunar bases. And a talking intelligent computer! Cool space suits! Video telephones!
Remember, this was before the first moon landing in 1969. There were Apollo flights happening, like Apollo 8 that orbited the moon in December 1968. There was a strong commitment to advancing space flight.
In 1968, the movie presented things that could actually happen, by the year 2001. Much more like a bunch of product announcements rather than science fiction (long ago in a galaxy far, far away).
Ages ago, the Digital Equipment VAX had a hardware hack where a null pointer (value is 0) would return a 0 when de-referenced. Which is the null character, the end of a string. So most of the time operations moving characters would terminate correctly. Obviously not portable to other computers.
In the IEEE 754 floating point definition, there are specific values for errors: mathematically undefined, division by zero, overflow, underflow, and inexact. These MUST be implemented in hardware to meet the specification.
Ages ago, before even dumb terminals, we were programming on punched cards, batch compiling, and printing on a line printer. For some reason, this printer supported control G (the bell character) with a buzzer. So someone hacked the compiler to include control G in the beginning of the error messages.
So we could sit back and listen to the line printer. If the buzzer sounded, the programmer would say "Oh &#$!@". Just like Pavlov demonstrated.
Google makes tons of money each quarter, but cancels projects / products, and lays off people. That does not improve moral. Why should I work my ass off and then my project is cancelled and I am laid off?
Problems in AI is leadership problems.
But what language is the program written in? It would have to include all of the possible data types and formats, and include all possible operations with the correct semantics.
Their climate calculations do not include the environmental costs of commuting and return to office mandates. There would be less pollution if no employee commuted and there is no big office building.
The Mythical Man Month by Frederick Brooks. The examples are old, but the wisdom remains. Particularly the chapter (and CACM article) No Silver Bullet-- Essence and Accident in Software Engineering (1986).
Or just: https://xkcd.com/1425/
System.out.println("Hello World");
Convince me that an AI Agent won't piss off the customer.
Some of his books are available in pdf on his website: https://gordonbell.azurewebsites.net/gbvita.htm
I used Computer Structures:Readings and Examples in an architecture class, and High Tech Ventures in an entrepreneurial class at Stanford.
I feel that Digital Equipment Corp entered a death spiral when he left.
As a volunteer, I was fortunate that I was able to work with him on some projects at the Computer History Museum. I wish that I had taped him discussing many of the artifacts in the exhibit, especially machines he worked on.
I created some training notebooks for CHM docents using chapters from Computer Structures, and the 1982 revised version.
There are a bunch of places in Silicon Valley. For example: https://www.abettersource.com/ Search for used office furniture.
There are also auctions when a company closes down: For example: https://svdisposition.com/auctions These may include furniture, computer equipment, machinery, lab equipment. Bidding may be on single items or many items batched together, like 50 office chairs.
There was a local auction two days ago (May 16, 2024) for the entire contents of a Red Lobster restaurant that suddenly closed, as one bid.
They are restricting the solutions by encapsulating each problem with boundaries.
1. Create remote work technology that is better than open plan workplaces.
2. Create small self-driving one or two person electric cars.
3. Rebuild the sprawling two story office parks with multistory housing: one desk => one apartment. Everyone can walk to work.
4. And housing on top of schools and shopping malls for staff.
5. Standardized modular homes that can be easily assembled into new buildings and slip into old office buildings.
6. Nasty filters for email, messages, social media posts and replies.
7. Formal verification for software, and hardware that protects it. Penalties for broken software.
There is a place in SF called luma that sponsors many such events. I have never been, don't know if it is a waste of time or not. https://lu.ma/sf
1. Winnie the Pooh to kids on their third or fourth or fifth birthday. 2. The Mythical Man Month to non-tech bosses. 3. The Moon is a Harsh Mistress to girlfriends (it's a test). 4. "Get It Together: Organize Your Records So Your Family Won't Have To" to 70+ old friend (single) and 70+ old sister (single).
Years ago, I ran a German article about an ex-vice president turned environmentalist through Google Translate. In the result his name was Aluminum Gore.
Is the generated code correct? Was the training input code correct? Did the AI understand what you were asking? Did you specify the problem correctly, with all of the boundary conditions?
If AI programming is so good, why are there so many zero-day bugs in Windows?
Install bollards in the yard in the path of the cars. These are vertical steel pipes set in concrete and filled with concrete. Paint them bright yellow. Correctly installed, they will stop a car with a big crunch. Probably cheaper than rebuilding a house wall again.
Move to San Francisco, or Silicon Valley and start networking: subscribe to techstars silicon valley sf bay area newsletter.
Triplebyte had online tests and phone interviews, now owned by Karat.com
Put a notification in the monthly 'whos hiring' thread on HN.
Put a posting on LinkedIn, Indeed, SimplyHired, Glassdoor, Dice, etc.
If your concept has any tech, you need a technical co founder to get it right. VCs will be interested in the tech experience of the founding team.
I tried to read the article, but received a popup in the face asking for a subscription. I guess the answer is, "Yea, a media extinction event has already happened".
You need to be able to generate any machine instruction. like clear the cache, turn interrupts off and on, access the hardware bits on the I/O cards, access privileged instructions, etc. The language needs to use every instruction. In C, there is a feature to embed/call assembly language instructions which are architecture dependent.
Want to prohibit links to sites with paywalls.
Go to the Stanford Business School. Learn to schmooze. Stanford has a golf course, with classes for students. Date and marry one of the students. Make lots of contacts. Hang out at the various startup groups and the D-school. Attend the biannual Y-Combinator Demo Days (Sept 6 and 7, 2023). Hang out with the VC's that circle the Bus school. Visit the Computer History Museum for inspiration.
Start with one or more of Harvard's intro CS courses: https://www.classcentral.com/report/harvard-cs50-guide/ Then you need a lot of data structure / algorithms work.
There are free online CS courses at Harvard, MIT, and Stanford. Computer Science is a huge field, with many specializations. Computer Science is the study of concepts, programming is the implementation.
Years ago, I watched a Y Combinator startup school video. The presenter was talking about their standard company structure, registering the company in Delaware, standard vesting schedule (four years, one year cliff, 1/48 per month).
One of the audience members interrupted, his co-founders had discussed their vesting schedule alot and thought it was better, he started to explain.
The presenter interrupted, asking why are you wasting time, energy, and brain cells on discussing vesting, and not spending it on THE PRODUCT, product market fit, MVP, customers, schedule, and THINGS THAT WILL MAKE YOUR company succeed. Vesting is binary, it either pays off, or more typically is worth zero. Don't waste the group's time and certainly don't waste VC's time on vesting. During your pitch, vesting should be exactly one sentence, the standard. Anything different makes you look clueless and not focusing on the PRODUCT.
There are a number of actual matchmaking services in San Francisco and Silicon Valley.
The wikipedia article on Garbage Collection (software) is one place to start: https://en.wikipedia.org/wiki/Garbage_collection_(computer_s...
Here is a list of garbage collection papers (newest is 2002) https://sites.cs.ucsb.edu/~ckrintz/racelab/gc/papers/
Here is an old paper (1994?) that surveys different kinds of software Garbage Collection. https://www.cs.cmu.edu/~fp/courses/15411-f08/misc/wilson94-g...
I am concerned that you are new to CS and you want to 'improve' a working system's garbage collection code. But good luck anyway. Finding a new fast GC is like searching for the Holy Grail.
680 units, only 212 parking spaces. Plus retail and gym that needs parking. On the geologic map, looks like the location is on sand (sand dunes). How much water, waste water treatment, electrical power will be required?