Aurora Driver has begun commercial operations. Aurora Driver is the eighteen-wheeler truck driving AI from Aurora Innovation. The first driverless truck delivery trip was on April 26th, between Dallas and Houston.
https://aurora.tech/newsroom/aurora-delivers-my-ride-in-the-first-self-driving-commercial-truck
What Mario Zechner learned these past months:
"Abandon Arduino and embrace the ESP32 line of microcontrollers, which are a million times more versatile, powerful and cheaper."
"Learn how to use Electronic Design Automation software like KiCad or in my case EasyEDA Pro, to create my own PCB designs"
"ApplyIQ: A smarter way to get hired."
"ApplyIQ is your personal AI job search agent. Upload your resume, tell us what you're looking for, and we apply to jobs for you."
So now employers can get swamped with thousands (millions?) of résumés, rather than mere hundreds? But hey, if you do it first you may get ahead of the curve and land that dream job?
A new benchmark called CURIE measures how good language models are at science.
https://research.google/blog/evaluating-progress-of-llms-on-scientific-problem-solving/
The CEO of Shopify has just announced that all developers must use AI daily. Actually all *employees*, developers or otherwise, must use AI daily.
https://x.com/tobi/status/1909251946235437514
#solidstatelife #ai #genai #llms #codingai #technologicalunemployment
"The Great Unstacking: How new US tariffs are forcing Europe to rethink its entire tech stack."
If you are in Europe, switch to: OVHcloud (France), IONOS (Germany), Scaleway (France), Elastx (Sweden), Hetzner (Germany).
"Vibe coding" is a security disaster.
"His business, built entirely with AI assistance and 'zero hand-written code,' was experiencing bypassed subscriptions, maxed-out API keys, and database corruption."
https://nmn.gl/blog/vibe-coding-fantasy
#solidstatelife #ai #genai #llms #codingai #vibecoding #cybersecurity
Manus AI is a new "agentic" AI from China that claims to exceed OpenAI's DeepResearch on a benchmark called "General AI Assistants" (GAIA).
New research on AI and code quality. The people who did the study last year showing AI increases "churn" did a follow-up study this year.
AI enables developers to produce new code substantially faster at the cost of long-term maintainability. Code is added about ~20% faster, but "churn" is ~50% higher, code duplication is ~30% higher, and refactorization is ~80% lower.
"Google is done. Compare the quality of these responses (ChatGPT)". Screenshots of search queries: "In LaTeX, how do I write a differential equation?" and "How do I mark a Solidity as callable only by the contract creator?"
#solidstatelife #ai #nlp #openai #gpt
twitter.com/jdjkelly/status/15…
"Shufflecake is a tool for Linux that allows creation of multiple hidden volumes on a storage device in such a way that it is very difficult, even under forensic inspection, to prove the existence of such volumes. Each volume is encrypted with a different secret key, scrambled across the empty space of an underlying existing storage medium, and indistinguishable from random noise when not decrypted. Even if the presence of the Shufflecake software itself cannot be hidden -- and hence the presence of secret volumes is suspected -- the number of volumes is also hidden. This allows a user to create a hierarchy of plausible deniability, where 'most hidden' secret volumes are buried under 'less hidden' decoy volumes, whose passwords can be surrendered under pressure. In other words, a user can plausibly 'lie' to a coercive adversary about the existence of hidden data, by providing a password that unlocks 'decoy' data. Every volume can be managed independently as a virtual block device, i.e. partitioned, formatted with any filesystem of choice, and mounted and dismounted like a normal disc. The whole system is very fast, with only a minor slowdown in I/O throughput compared to a bare LUKS-encrypted disk, and with negligible waste of memory and disc space."
"You can consider Shufflecake a 'spiritual successor' of tools such as Truecrypt and Veracrypt, but vastly improved. First of all, it works natively on Linux, it supports any filesystem of choice, and can manage up to 15 nested volumes per device, so to make deniability of the existence of these partitions really plausible."
Introducing Shufflecake: plausible deniability for multiple hidden filesystems on Linux
#solidstatelife #cryptography #steganography
"The world's first programming language based on classical Chinese is only about a month old, and volunteers have already written dozens of programs with it, such as one based on an ancient Chinese fortune-telling algorithm."
The core of the language "includes a renderer that can display a program in a manner that resembles pages from ancient Chinese texts."
I lucked out that the language used for programming languages is my native language, English.
"Currently wenyan-lang contributors are working on transpilers for Python, Ruby, JavaScript, C++, and Java, libraries for graphics and the graphical user interface (GUI)" and "Lingdong Huang is currently working on an introductory guide to programming in wenyan-lang that is itself written in classical Chinese."
World's First Classical Chinese Programming Language
#solidstatelife #computerscience #programminglanguages #chinese
The rise and fall of Bitcoin maximalism. Has Bitcoin maximalism fallen?
"I trace the origin of today's maximalism back to the Bitcoin Civil War. As we've seen, the term originates in 2014, but truly takes off after following a contentious debate about blocksize and scaling solutions that led to the Bitcoin Cash hard fork in 2017. A particular vision for Bitcoin won out -- favoring smaller blocks, SegWit implementation and scaling through the Lightning Network."
"Instead of Bitcoin primarily as digital cash with some gold-like properties, as envisioned by Roger Ver, Bitcoin would be primarily digital gold at the base layer, but with a dedicated layer 2 for cash-like payments."
"In hindsight, the Dragon's Den set the stage for maximalism's emphasis on adversarial thinking -- it is Bitcoin against everyone - where attack is the best defence."
"Dragon's Den" refers to a bitcoin Slack channel.
"The following day, large blockers discovered a video presentation by Bram Cohen, the inventor of Bittorrent, which was given in January 2017. During the presentation, a Slack channel called 'dragonsden' accidentally popped up on screen, a channel on the Bitcoin Core Slack. It was a private channel that had 21 members. On the screen, it was possible to identify a few members, who the large blockers perceived as well-known small block trolls, including some of the moderators of the Bitcoin subreddit. To the large blockers, this was a major scandal, evidence of coordination on the small block side, a link between the Bitcoin subreddit, Bitcoin Core and a propaganda campaign. The large blockers are also likely to have had secret communication and coordination channels focusing on public relations. This was the political nature of the situation: the conflict had evolved to an extent where such communication channels were necessary. To me, there was one clear takeaway from this 'scandal'; I simply had to get inside the Dragons' Den. A few weeks later, after asking around a bit, I was in! I made it into the depths of the Dragons' Den."
"The channel was very active and entirely focused on the blocksize war. Most of the conversion revolved around social media, public relations and how best to expose weaknesses in the narratives and arguments put forward by the large blockers. Many participants in the channel appeared very committed to the cause. Discussion often focused around how to convince various people to join the small block camp, who was likely to turn, and which would be the most effective issues to focus on in social media. There was also discussion on memes and meme production. This war was also a meme war, and the 'Dragons', as members of the channel were sometimes called, were very involved in meme production. Many of these memes were humorous, designed to make the large blockers appear as if they had a weak understanding of some of the technical issues in Bitcoin."
The rise and fall of Bitcoin culture
#solidstatelife #cryptocurrency #bitcoin
The rise and fall of Bitcoin maximalism. Has Bitcoin maximalism fallen?
"I trace the origin of today's maximalism back to the Bitcoin Civil War. As we've seen, the term originates in 2014, but truly takes off after following a contentious debate about blocksize and scaling solutions that led to the Bitcoin Cash hard fork in 2017. A particular vision for Bitcoin won out -- favoring smaller blocks, SegWit implementation and scaling through the Lightning Network."
"Instead of Bitcoin primarily as digital cash with some gold-like properties, as envisioned by Roger Ver, Bitcoin would be primarily digital gold at the base layer, but with a dedicated layer 2 for cash-like payments."
"In hindsight, the Dragon's Den set the stage for maximalism's emphasis on adversarial thinking -- it is Bitcoin against everyone - where attack is the best defence."
"Dragon's Den" refers to a bitcoin Slack channel.
"The following day, large blockers discovered a video presentation by Bram Cohen, the inventor of Bittorrent, which was given in January 2017. During the presentation, a Slack channel called 'dragonsden' accidentally popped up on screen, a channel on the Bitcoin Core Slack. It was a private channel that had 21 members. On the screen, it was possible to identify a few members, who the large blockers perceived as well-known small block trolls, including some of the moderators of the Bitcoin subreddit. To the large blockers, this was a major scandal, evidence of coordination on the small block side, a link between the Bitcoin subreddit, Bitcoin Core and a propaganda campaign. The large blockers are also likely to have had secret communication and coordination channels focusing on public relations. This was the political nature of the situation: the conflict had evolved to an extent where such communication channels were necessary. To me, there was one clear takeaway from this 'scandal'; I simply had to get inside the Dragons' Den. A few weeks later, after asking around a bit, I was in! I made it into the depths of the Dragons' Den."
"The channel was very active and entirely focused on the blocksize war. Most of the conversion revolved around social media, public relations and how best to expose weaknesses in the narratives and arguments put forward by the large blockers. Many participants in the channel appeared very committed to the cause. Discussion often focused around how to convince various people to join the small block camp, who was likely to turn, and which would be the most effective issues to focus on in social media. There was also discussion on memes and meme production. This war was also a meme war, and the 'Dragons', as members of the channel were sometimes called, were very involved in meme production. Many of these memes were humorous, designed to make the large blockers appear as if they had a weak understanding of some of the technical issues in Bitcoin."
The rise and fall of Bitcoin culture
#solidstatelife #cryptocurrency #bitcoin
OpenStreetMap 2012 vs 2022. Side-by-side comparison centered around New York. Much more detail in the 2022 version.
OpenStreetMap 2012 vs 2022
#solidstatelife #cartography #openstreetmap
"Ultra-large AI models are over." "I don't mean 'over' as in 'you won't see a new large AI model ever again' but as in 'AI companies have reasons to not pursue them as a core research goal -- indefinitely.'" "The end of 'scale is all you need' is near."
He (Alberto Romero) breaks it down into technical reasons, scientific reasons, philosophical reasons, sociopolitical reasons, and economic reasons. Under technical reasons he's got new scaling laws, prompt engineering limitations, suboptimal training settings, and unsuitable hardware. Under scientific reasons, he's got biological neurons vastly greater than artificial neurons, dubious construct validity and reliability, the world is multimodal, and the AI art revolution. Under philosophical reasons he's got what is AGI anyway, human cognitive limits, existential risks, and aligned AI, how? Under sociopolitical reasons he's got the open-source revolution, the dark side of large language models, and bad for the climate. Under economic reasons, he's got the benefit-cost ratio is low and good-enough models.
Personally, I find the "scientific reasons" most persuasive. I've been saying for a long time that we keep discovering the brain is more common than previously thought. If that's true, it makes sense that there are undiscovered algorithms for intelligence we still need in order to make machine intelligence comparable to human intelligence. If the estimates here that to simulate biological dendrites, you need hundreds of artificial neurons, and to simulate a whole biological neuron, you need a thousand or so artificial neurons, that fits well with that picture.
Having said that, the gains recently from simple scaling up the size of the large language models have been impressive. Having said that, notice that in the visual domain, it's been algorithmic breakthroughs, in this case what are known as diffusion networks, that have driven recent progress.
Ultra-large AI models are over
#solidstatelife #ai #openai #gpt3 #llms #agi
"When we navigate in a new environment, we are required to pay attention to our surroundings and to update our position using our own internal navigation system in order to reach our destination. Using GPS removes these requirements and renders navigation less cognitively demanding. In fact, people who travel along given routes using GPS gain less knowledge about those routes compared to people who travel the same routes without an aid, using a map, or after being guided by an experimenter. However, no studies have looked at whether GPS use has long-term effects on our internal navigation system, when we are required to find our way without a navigation aid."
"When we navigate without GPS in a new environment, there are two navigation strategies that we can use that depend on separate brain systems. One is the spatial memory strategy and involves learning the relative positions of landmarks and serves to form a cognitive map of the environment. This strategy critically relies on the hippocampus, a brain region heavily involved in episodic memory and relational memory. The other strategy is the stimulus-response strategy and involves learning a sequence of motor responses (e.g., turn left) from specific positions (e.g., next corner). Stimulus-response learning critically relies on the caudate nucleus, a brain region also responsible for habit learning (e.g., learning how to ride a bicycle). This strategy leads to more rigid behavior and allows us to navigate on 'auto-pilot' on routes that we travel frequently."
"Using GPS involves following step-by-step sensorimotor instructions, which is similar to learning stimulus-response associations (e.g., turn right at the next intersection, turn left in 500 m). In a cross-sectional study, we sought to determine whether individuals with greater GPS habits rely more on stimulus-response strategies and less on spatial memory strategies when they are required to navigate without GPS, and whether they have poorer cognitive mapping abilities and landmark encoding."
"There was no significant correlation between lifetime GPS experience and subjective sense-of-direction scores. Thus, it does not appear that participants who used GPS for more hours did so as a result of a subjectively poor sense of direction."
"We found a significant negative correlation between lifetime GPS experience (in hours) and performance on the first probe trial of the concurrent spatial discrimination learning task, and a marginally significant negative correlation with performance on both probe trials combined, indicating that people with more lifetime GPS experience use hippocampus-dependent spatial memory strategies to a lesser extent."
"With regards to GPS reliance, there was a significant negative correlation with both probe trials combined, which indicates that as GPS reliance increases, spatial memory strategy use decreases." "In terms of sense of GPS dependence, there was a significant positive correlation between sense of GPS dependence scores and the second probe trial of the concurrent spatial discrimination learning task. This indicates that those who feel more dependent on their GPS have a lower ability to learn from their mistakes in the first probe trial, regardless of their subjective sense of direction. There was also a significant correlation with both probe trials combined, indicating a lower use of spatial memory strategies in those who feel more dependent on their GPS."
"We also examined the question of whether perceived spatial abilities are related to actual navigational performance on the concurrent spatial discrimination learning task. Subjective sense-of-direction scores were significantly related to faster learning, and, surprisingly, to lower performance on the second probe trial, and on both trials combined."
On the 4-on-8 virtual maze, "We found a significant negative correlation between lifetime GPS experience and navigation strategy scores. This indicates that individuals with more lifetime GPS experience use hippocampus-dependent spatial memory strategies to a lesser extent, concordant with the concurrent spatial discrimination learning task probe results. People with greater lifetime GPS experience also had more difficulty forming a cognitive map, as evidenced by a significant negative correlation between lifetime GPS experience and map drawing scores. This can at least in part be explained by the fact that people with more lifetime GPS experience encoded fewer landmarks, as there were significant negative correlations between lifetime GPS experience and the average number of landmarks used while solving the task, as well as the number of landmarks they noticed in the environment."
Habitual use of GPS negatively impacts spatial memory during self-guided navigation
#neuroscience #solidstatelife #gps
Movie made entirely by Google's image-generating AI, Imogen. Very short movie, about 2 minutes. It's called "Tiger Goes to the City" and it's about a tiger who wants some friends, so goes to the city looking for some friends.
Have we reached the era where anybody can type some words into a computer and make a movie? Seems like yes although it's kind of a whacky movie, and has no sound so it's like from the silent movie era.
https://www.youtube.com/watch?v=wONzm6eIBAY
#solidstatelife #ai #computervision #generativeai
DALL-E-2 open source clone under development. "Startup Stability AI now announced the release of Stable Diffusion, another DALL-E 2-like system that will initially be gradually made available to new researchers and other groups via a Discord server."
"After a testing phase, Stable Diffusion will then be released for free -- the code and a trained model will be published as open source. There will also be a hosted version with a web interface for users to test the system."
"Stable Diffusion is the result of a collaboration between researchers at Stability AI, RunwayML, Ludwig Maximilian University of Munich (LMU Munich), EleutherAI and Large-scale Artificial Intelligence Open Network (LAION)."
Open-source rival for OpenAI’s DALL-E runs on your graphics card
#solidstatelife #ai #computervision #diffusionnetworks #openai #dalle #opensource
"SPAR19 becomes the most tracked flight of all time on Flightradar24." "From the time it lifted off from Kuala Lumpur at 15:42 local time, SPAR19 was already the most tracked flight on Flightradar24 among active flights. By the time it landed in Taipei, SPAR19 was being tracked by more than 708,000 people around the world, making it the most tracked live flight in Flightradar24 history. Over the seven hours from Kuala Lumpur to Taipei, a total of 2.92 million people followed at least a portion of the flight."
So, we live in a time when millions of people can follow a single flight online. In case you're wondering what was so special about that flight, one... er, two words, but one proper noun: Nancy Pelosi.
SPAR19 becomes the most tracked flight of all time
#solidstatelife #aviation
