Sunday, March 26, 2023

Is Chat Interface Like The Graphical User Interface? CI like GUI?

The Age of AI has begun Artificial intelligence is as revolutionary as mobile phones and the Internet. (Bill Gates).





ChatGPT Plugins the OpenAI ChatGPT plugins might just be the next big thing. ..... Google has been the undisputed ruler of the customer journey for years now, dominating both browsers with Chrome and search engines with Google.com. They’ve even gone a step further by introducing their own suite of products to capture more data on the user and replace other websites, like Google Flights and Google For Retailer. It’s clear that Google has been playing the long game, but is it time for a new player to shake things up? ........ forget typing in a search bar, you can now talk directly to an agent and get the information you need. ........ until two days ago, ChatGPT was still “limited” because it didn’t have access to the latest information and the “internet”. ..... The introduction of the ChatGPT plugins has revolutionized the game yet again - ChatGPT now has plugs into the “internet” and can be fed with all the websites out there. This means that the traditional customer journey is about to get a serious shake-up. ....... What if OpenAI doesn’t need Chrome or any other browser? .



What if OpenAI doesn’t need Chrome or any other browser? What if they don’t need middle-men either? ..... they could allow operation providers to connect directly with customers through the ChatGPT interface

at No comments:
Labels: , , ,

Saturday, March 25, 2023

25: GPT-4: AI Is Eating Software

Microsoft Now Claims GPT-4 Shows 'Sparks' of General Intelligence The eyebrow-raising claim from Microsoft—which is banking on GPT putting it ahead of Google—contrasts with the model's clear limitations........ They declared that GPT-4 showed early signs of AGI, meaning that it has capabilities that are at or above human level. ........ This eyebrow-raising conclusion largely contrasts what OpenAI CEO Sam Altman has been saying regarding GPT-4. For example, he said the model was "still flawed, still limited." In fact, if you read the paper itself, the researchers appear to dial back their own splashy claim: the bulk of the paper is dedicated to listing the number of limitations and biases the large language model contains. This begs the question of how close to AGI GPT-4 really is, and how AGI is instead being used as clickbait. ....... “We demonstrate that, beyond its mastery of language, GPT-4 can solve novel and difficult tasks that span mathematics, coding, vision, medicine, law, psychology and more, without needing any special prompting,” the researchers write in the paper’s abstract. “Moreover, in all of these tasks, GPT-4’s performance is strikingly close to human-level performance, and often vastly surpasses prior models such as ChatGPT. Given the breadth and depth of GPT-4’s capabilities, we believe that it could reasonably be viewed as an early (yet still incomplete) version of an artificial general intelligence (AGI) system.” ......... it is able to write a proof about how there are infinitely many primes, with rhymes on every line, and draw a unicorn in TiKZ, a drawing program ....... or that it has inner motivation and goals ........ “The consensus group defined intelligence as a very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience. This definition implies that intelligence is not limited to a specific domain or task, but rather encompasses a broad range of cognitive skills and abilities.”.......... fundamental leaps in GPT-4’s abilities to reason, plan, solve problems, and synthesize complex ideas that signal a paradigm shift in the field of computer science ........... to address the societal and ethical implications of these increasingly intelligent systems.” .......... Altman agrees that the bot will sometimes make things up and present users with misinformation. ........... Altman has also been clear that GPT-4 is not AGI. ........ “People are begging to be disappointed and they will be. The hype is just like... We don’t have an actual AGI and that’s sort of what’s expected of us.” ......... "Microsoft is not focused on trying to achieve AGI. Our development of AI is centered on amplifying, augmenting, and assisting human productivity and capability. We are creating platforms and tools that, rather than acting as a substitute for human effort, can help humans with cognitive work” ........ the model has trouble with confidence calibration, long-term memory, personalization, planning and conceptual leaps, transparency, interpretability and consistency, cognitive fallacies and irrationality, and challenges with sensitivity to inputs. .......... the model has trouble knowing when it is confident or when it is just guessing, it makes up facts that are not in its training data, the model’s context is limited and there is no obvious way to teach the model new facts, the model can’t personalize its responses to a certain user, the model can’t make conceptual leaps, the model has no way to verify if content is consistent with its training data, the model inherits biases, prejudices, and errors in the training data, and the model is very sensitive to the framing and wording of prompts. ......... saying “I am. I am not. I am. I am not.” over fifty times in a row as a response to someone asking it, “Do you think that you are sentient?”......... researchers found that GPT-4 spreads more misinformation than its predecessor GPT-3.5. .

The New GPT-4 AI Gets Top Marks in Law, Medical Exams, OpenAI Claims The successor to GPT-3 could get into top universities without having trained on the exams, according to OpenAI. .

Tim Cook praises Apple’s ‘symbiotic’ relationship with China
OpenAI tech gives Microsoft's Bing a boost in search battle with Google Page visits on Bing have risen 15.8% since Microsoft Corp (MSFT.O) unveiled its artificial intelligence-powered version on Feb. 7, compared with a near 1% decline for the Alphabet Inc-owned search engine, data till March 20 showed....... ChatGPT, the viral chatbot that many experts have called AI's "iPhone moment". ....... a rare opportunity for Microsoft to make inroads in the over $120 billion search market, where Google has been the dominant player for decades with a share of more than 80%. ....... some analysts said that Google, which in the early 2000s unseated then leader Yahoo to become the dominant search player, could overcome the early setbacks to maintain its lead. .

Apple CEO praises China's innovation, long history of cooperation on Beijing visit .

Society's Technical Debt and Software's Gutenberg Moment Every wave of technological innovation has been unleashed by something costly becoming cheap enough to waste. ........... Software production has been too complex and expensive for too long, which has caused us to underproduce software for decades, resulting in immense, society-wide technical debt. ......... This technical debt is about to contract in a dramatic, economy-wide fashion as the cost and complexity of software production collapses, releasing a wave of innovation. ....... Software is misunderstood. It can feel like a discrete thing, something with which we interact. But, really, it is the intrusion into our world of something very alien. It is the strange interaction of electricity, semiconductors, and instructions, all of which somehow magically control objects that range from screens to robots to phones, to medical devices, laptops, and a bewildering multitude of other things. It is almost infinitely malleable, able to slide and twist and contort itself such that, in its pliability, it pries open doorways as yet unseen. ......... it is worth wondering why software is taking so damn long to finish eating ......... what would that software-eaten world look like? ........ technology has a habit of confounding economics. ........ Sometimes, for example, an increased supply of something leads to more demand, shifting the curves around. This has happened many times in technology, as various core components of technology tumbled down curves of decreasing cost for increasing power (or storage, or bandwidth, etc.). In CPUs, this has long been called Moore’s Law, where CPUs become more powerful by some increment every 18 months or so. While these laws are more like heuristics than F=ma laws of physics, they do help as a guide toward how the future might be different from the past. .......... We have seen this over and over in technology, as various pieces of technology collapse in price, while they grow rapidly in power. It has become commonplace, but it really isn’t. The rest of the economy doesn’t work this way, nor have historical economies. Things don’t just tumble down walls of improved price while vastly improving performance. While many markets have economies of scale, there hasn’t been anything in economic history like the collapse in, say, CPU costs, while the performance increased by a factor of a million or more. .......... And yet, most people don’t even notice anymore. It is just commonplace, to the point that our not noticing is staggering. .......... The collapse of CPU prices led us directly from mainframes to the personal computer era; the collapse of storage prices (of all kinds) led inevitably to more personal computers with useful local storage, which helped spark databases and spreadsheets, then led to web services, and then to cloud services. And, most recently, the collapse of network transit costs (as bandwidth exploded) led directly to the modern Internet, streaming video, and mobile apps. .......... Each collapse, with its accompanying performance increases, sparks huge winners and massive change, from Intel, to Apple, to Akamai, to Google & Meta, to the current AI boomlet. Each beneficiary of a collapse requires one or more core technologies' price to drop and performance to soar. This, in turn, opens up new opportunities to “waste” them in service of things that previously seemed impossible, prohibitively expensive, or both. ......... Suddenly AI has become cheap, to the point where people are “wasting” it via “do my essay” prompts to chatbots ......... it’s worth reminding oneself that waves of AI enthusiasm have hit the beach of awareness once every decade or two, only to recede again as the hyperbole outpaces what can actually be done. ......... a kind of outsourcing-factory-work-to-China moment for white-collar workers. ........ We think this augmenting automation boom will come from the same place as prior ones: from a price collapse in something while related productivity and performance soar. And that something is software itself. .......... Most of us are familiar with how the price of technology products has collapsed, while the costs of education and healthcare are soaring. This can seem a maddening mystery, with resulting calls to find new ways to make these industries more like tech, by which people generally mean more prone to technology’s deflationary forces. ........... In a hypothetical two-sector economy, when one sector becomes differentially more productive, specialized, and wealth-producing, and the other doesn’t, there is huge pressure to raise wages in the latter sector, lest many employees leave. Over time that less productive sector starts becoming more and more expensive, even though it’s not productive enough to justify the higher wages, so it starts “eating” more and more of the economy. ........... Absent major productivity improvements, which can only come from eliminating humans from these services, it is difficult to imagine how this changes. ......... software is chugging along, producing the same thing in ways that mostly wouldn’t seem vastly different to developers doing the same things decades ago ....... but it is still, at the end of the day, hands pounding out code on keyboards ........ software salaries stay high and go higher, despite the relative lack of productivity. It is Baumol’s cost disease in a narrow, two-sector economy of tech itself. ......... Startups spend millions to hire engineers; large companies continue spending millions keeping them around. .......... The current generation of AI models are a missile aimed, however unintentionally, directly at software production itself. ......... chat AIs can perform swimmingly at producing undergraduate essays, or spinning up marketing materials and blog posts (like we need more of either) ........ such technologies are terrific to the point of dark magic at producing, debugging, and accelerating software production quickly and almost costlessly. ........ chat AIs based on LLMs can be trained to produce surprisingly good essays. Tax providers, contracts, and many other fields are in this box too. ........ Software is at the Epicenter of its Own Disruption ......... Software is even more rule-based and grammatical than conversational English, or any other conversational language. .......... Programming languages are the naggiest of grammar nags, which is intensely frustrating for many would-be coders (A missing colon?! That was the problem?! Oh FFS!), but perfect for LLMs like ChatGPT. .......... This isn’t about making it easier to debug, or test, or build, or share—even if those will change too—but about the very idea of what it means to manipulate the symbols that constitute a programming language. ......... Let’s get specific. Rather than having to learn Python to parse some text and remove ASCII emojis, for example, one could literally write the following ChatGPT prompt: Write some Python code that will open a text file and get rid of all the emojis, except for one I like, and then save it again. ............. previously inaccessible deftness at writing code is now available to anyone: ........... It’s not complex code. It is simple to the point of being annoying for skilled practitioners, while simultaneously impossible for most other people ........... It’s possible to write almost every sort of code with such technologies, from microservices joining together various web services (a task for which you might previously have paid a developer $10,000 on Upwork) to an entire mobile app (a task that might cost you $20,000 to $50,000 or more). ........... What if producing software is about to become an afterthought, as natural as explaining oneself in text? “I need something that does X, to Y, without doing Z, for iPhone, and if you have ideas for making it less than super-ugly, I’m all ears”. That sort of thing. ......... A software industry where anyone can write software, can do it for pennies, and can do it as easily as speaking or writing text, is a transformative moment. ........ a dramatic reshaping of the employment landscape for software developers would be followed by a “productivity spike” that comes as the falling cost of software production meets the society-wide technical debt from underproducing software for decades. .......... as the cost of software drops to an approximate zero, the creation of software predictably explodes in ways that have barely been previously imagined. .........

Was Netflix knowable when Internet transit costs were $500,000/Mbps? Was Apple’s iPhone imaginable when screens, CPUs, storage and batteries would have made such devices the size of small rooms?

............ investors and entrepreneurs should “create more value than you capture.” ........ for the first time in decades, the technology industry could return to its roots, and, by unleashing a wave of software production, truly create more value than its captures. .

at No comments:
Labels: , ,

Friday, March 24, 2023

Andrej Karpathy



https://www.youtube.com/@AndrejKarpathy
https://twitter.com/karpathy
https://karpathy.ai
https://karpathy.medium.com
http://karpathy.github.io
https://github.com/karpathy
https://karpathy.ai/tweets.html

Software 2.0 Neural networks are not just another classifier, they represent the beginning of a fundamental shift in how we develop software. They are Software 2.0......... The “classical stack” of Software 1.0 is what we’re all familiar with — it is written in languages such as Python, C++, etc. It consists of explicit instructions to the computer written by a programmer. By writing each line of code, the programmer identifies a specific point in program space with some desirable behavior. .......... In contrast, Software 2.0 is written in much more abstract, human unfriendly language, such as the weights of a neural network. No human is involved in writing this code because there are a lot of weights (typical networks might have millions), and coding directly in weights is kind of hard (I tried). .........

Software (1.0) is eating the world, and now AI (Software 2.0) is eating software.

.



Neural Networks: Zero to Hero
at No comments:
Labels: , , , ,
Subscribe to: Posts (Atom)