I saw a live demo of this recently and it was very impressive. It is very difficult to search the govt knowledge base because sources are in so many formats, different location, and have different hoops to gain access not to mention code words, acronyms, abbreviations, and the mixing of technical language with contract language. I think it will mostly used as a search tool rather than generating ideas.
I think the worry about GPT-4 making things up is valid. We all know what happened to they lawyer who used GPT. But I think this comes with training. Users need to be trained to use it as tool and to verify the outputs. Now will everyone do this? No. There are lazy and incompetent people in every large organization and the govt is no different.
The concern about a LLM influencing or biasing its output is also a worry. Maybe there isn't a good solution to this one. I would say that having a govt group testing and assessing it would be best however they don't have the expertise form such a group which is the whole reason why they are leaning on companies like MS to guide their AI usage in the first place. I could also argue that this may not matter much anyway since the govt decisions are already heavily influenced by lobbyist and illegal promotion / favoritism of contractors.
Near as I can tell the only graphical game released for the Tandy 2000 was a special edition of Microsoft Flight Simulator that also supported the Tandy 1000 and 1200HD. The 2000 was in their business line, not quite PC compatible but it ran DOS and was supposed to run applications like MultiMate, dBASE, and Basic Four.
The Tandy 1000 line did support Sierra games very well.
> The Tandy 1000 line did support Sierra games very well.
Very true! I wonder if the Sierra games (e.g. King's Quest, Space Quest) were some of the "killer apps" for Tandy computers, i.e. the apps which made people want to buy a Tandy computer. Radio Shack certainly sold Sierra games and I always wanted the latest one for Christmas.
In the business world, things like spreadsheets were the killer apps, but I certainly didn't care about spreadsheets as a kid.
> Very true! I wonder if the Sierra games (e.g. King's Quest, Space Quest) were some of the "killer apps" for Tandy computers, i.e. the apps which made people want to buy a Tandy computer.
Before VGA and Sound Blaster came along, almost certainly. The Tandy 1000 line supported 3-voice sound and more colorful graphics in a manner almost identical to the PCjr, just in a less jank package. So they looked and sounded better on a Tandy than on most contemporary 80s PCs. Sierra games not only were sold at Radio Shack, but Tandy cross-promoted them in their store, even to the point of running special Sierra/Tandy demos on in-store machines to attract buyers.
I see other people on here say this but I think it is so important that it can be said again. WRITE, WRITE, WRITE! As you read and learn I think it is so important to write. I suggest a mixture of notes, journal entries, and short essays. The more you do this the more your thoughts will gain clarity and you will see themes emerge from what you are doing.
As far as finding 'what to do with it'. I think that will develop over time as you talk with people about your interests. Share your writings as best as you can and join discussions. See if there are local MeetUp (or something similar) groups on the topics that you find most interesting. If so, join them and go to discussions. Ask if you can give short presentations on topics to the group. If you find a book or essay particularly interesting then send an email to the author.
Another trick that has worked for me; Some famous authors can be sometimes really hard to reach. But look at the work that they cite. Usually you will find some academic paper that they quote or reference. Look up the author / authors of those papers and read about them and contact them directly. Professors LOVE to talk about their work and usually (from my experience) they are happy to talk to people about their research who have read their papers. From those conversations you might find a path to a career or hobby that make you feel fullfilled. Enjoy the journey.
How do you prevent this from turning into another FB? Is there a difference between using this and using FB where you have all of your posts / activity only viewable with your friends? I guess one obvious difference is FB's advertising and data mining. Are there any features that you absolutely will not add to Circles in the future because they will 'cross the line' into turning Circles into a FB clone?
It's a very valid question, the answer is that I don't think the market wants another FB (especially after today). In that, and from continually trying to scratch my own itch, we need to build something that works in an entirely different way.
If you try the app, we've turned a lot of 'standard' social features on their head, like how interactions work and are viewed in the app. But to answer your question:
- We will not ask for any personal information (such as job, school, interests etc).
- We will not add noise on the feed, such as what other people are liking/commenting on.
- We will not have any third party content on the feed. Eg anything that isn't from your contacts you have circled.
- We currently (and will continue) to only show you posts from contacts you have actively circled.
- We currently (and will continue to) keep the feed in chronological order.
I really agree with your comments; especially number 1. Often I can't use some black box implementation of an solver (or other algorithm for that matter) without some modifications. Numerical stability is a big one but also just performance. Sometimes the mathematically correct way of doing something is not always the best in practice. Short cuts and approximations can provide huge benefits. It is difficult to make those modifications without understanding the inner workings of the original method.
One example is that GJ stops being generally efficient in arbitrary precision settings. Many people will never have to deal with this, but if you are doing cryptography, it matters.
For example, while you can use GJ to calculate the determinant of a matrix, this can easily become exponential, and for integer matrices (or generally for matrices over division rings), there is an alternative method (the Bareiss algorithm) that is actually efficient.
I had an issue with my pixel and had to send it in. The phone got ran over by a car and was smashed. It could be turned on but I could not do a factory reset. I have financial and other private information on it. So the first thing I did was try to keep it on for days until the battery ran out to see if the hardware encryption would help secure the data. After that I still wasn't sure if the information was secure enough so I took a drill through it to destroy as many of the ICs that I could. My wife then pointed out that maybe this self destruction that I did would violate some terms and void the insurance even though the phone was useless after being run over by the car. So, in the end, I decided to just buy a new phone without doing a trade in or insurance claim. The $500 or so was worth not risking PI.
remember in many part of world $150 is like 1 month income. People might have privilege to destroy phone but 90% of people are simply going to repair it.
I was never a big social media person. But back in 2012 I totally went off of social media all together. It happened because I was in grad school and my girlfriend at the time had lots of anxiety issues. Her doctor recommended that she delete her facebook. She was worried about "missing out" on things if she deleted it so her and I decided to delete it together at the same time. 10 years have passed and I still don't have an account and don't miss it at all! The only social media that I have is LinkedIn (and HN) for work purposes but haven't logged into it more than a few times a year.
There are some challenges to not having social media however. It has been integrated into so many other apps and platforms that it is hard to get away from. For example, some of my close friends have a Messenger group chat. Sometimes I have missed out on planning events because of that. Also platforms like Oculus started to require a FB to log in. Online dating was also hard. Dates would ask for my FB or Instagram account and when I told them that I didn't have one they would think I was trying to hide something. I don't want to totally bash social media BUT I really hope that it goes back to what it was originally made for and all of the integrations stop. I don't think this is the trend though...
Good question. Also consider that most of the threats to GPS are ground based jamming. It is fairly easy and cheap to make a jammer out of hobbyist parts and knock out a large area of GPS coverage. However, just because someone jams GPS for an entire city doesn't mean that they are able to jam the LEO satellites flying over that city.
SpaceX will (and does) have good coverage but the accuracy will not be to anywhere what people have come to expect from GPS. At least not without major design changes to focus on this as a product from SpaceX.
Yotta Navigation | Onsite / Remote | Full Stack Developer, Sensor Fusion / ML / Navigation Engineer, UI / UX Engineer, C/C++ Developer, Technical Program Manager | Full-time | Santa Clara, CA | yottanav.com/jobs
Yotta Nav. is an exciting small engineering company started in 2008 to create novel navigation hardware and algorithms. We specialize in sensor fusion and machine learning for wearable and handheld navigation devices. We are currently looking to fill two key senior level roles; a full-stack engineer for internal tools and a sensor fusion / machine learning engineer. HOWEVER we are looking for people for all roles listed above. In general, we are looking for some great developers. If you are interested and you don't match any of the listed job postings please feel free to still contact us.
Submit an application at yottanav.com/jobs or email jobs@yottanav.com
I think the worry about GPT-4 making things up is valid. We all know what happened to they lawyer who used GPT. But I think this comes with training. Users need to be trained to use it as tool and to verify the outputs. Now will everyone do this? No. There are lazy and incompetent people in every large organization and the govt is no different.
The concern about a LLM influencing or biasing its output is also a worry. Maybe there isn't a good solution to this one. I would say that having a govt group testing and assessing it would be best however they don't have the expertise form such a group which is the whole reason why they are leaning on companies like MS to guide their AI usage in the first place. I could also argue that this may not matter much anyway since the govt decisions are already heavily influenced by lobbyist and illegal promotion / favoritism of contractors.