nmaley

" Do we have any standard benchmarks for humanoids to do domestic tasks?" The answer is yes. Steve Wozniak proposed the Coffee Test. See https://www.youtube.com/watch?v=MowergwQR5Y

It's actually very clever. Despite the apparent simplicity, no current model could pass it.

Re your forecasts, I think they are optimistic in terms of timing but not ridiculously so.

I use Claude. It's really good, but you should try to use it as Boris suggests. The other thing I do is give it very careful and precisely worded specs for what you want it to do. I have the habit, born from long experience, of never assuming that junior programmers will know what you want the program to do unless you make it explicit. Claude is the same. LLM code generators are terrific, but they can't second guess unclear communication.

Using carefully written specs, I've found Claude will produce flawless code for quite complex problems. It's magic.

LLMs are parameter based representations of linguistic representations of the world. Relative to robot predictive control problems, they are low dimensional and static. They are batch trained using supervised learning and are not designed to manage real time shifts in the external world or the reward space. They work because they operate in abstract, rule governed spaces like language and mathematics. They are ill suited to predictive control tasks. They are the IBM 360s of AI. Even so, they are astonishing achievements.

LeCun is right to say that continuous self supervised (hierarchical) learning is the next frontier, and that means we need world models. I'm not sure that JEPA is the right tool to get us past that frontier, but at the moment there are not a lot of alternatives on the table.

I've spent a lot of my career doing various types of solution design. One of the insights I gained from thinking a lot about representation, intentionality and the philosophy of language is that the way you represent a problem has a big influence on how easy you will find it to solve the problem. I've found that helps with solution design. Don't just think about the problem. Think about what is the best way to represent the problem.

Majored in Philosophy. Started programming in 1973 on mainframes. Became a full time developer, systems analyst. 72 years old now, with 50 years experience in IT. Co-founded a couple of start ups, made a little bit of money. Went back to corporate life for a while. Ended up as a Program Architect at Salesforce. Resigned to start a company which develops and delivers commercial LLM/RAG solutions. Going reasonably well. Simple principles: keep learning, do what you want to do, not just what the man tells you. I saw a note from another Philosophy grad saying that Philosophy is actually useful in that it gives you a framework and a perspective to look at things a little differently. I agree with that.

I'm in the process of actually building LLM based apps at the moment, and Martin Fowler's comments are on the money. The fact is seemingly insignificant changes to prompts can yield dramatically different outcomes, and the odd new outcomes have all these unpredictable downstream impacts. After working with deterministic systems most of my career it requires a different mindset.

It's also a huge barrier to adoption by mainstream businesses, which are used to working to unambiguous business rules. If it's tricky for us developers it's even more frustrating to end users. Very often they end up just saying, f* it, this is too hard.

I also use LLM's to write code and for that they are a huge productivity boon. Just remember to test! But I'm noticing that use of LLM's in mainstream business applications lags the hype quite a bit. They are touted as panaceas, but like any IT technology they are tricky to implement. People always underestimate the effort necessary to get a real return, even with deterministic apps. With indeterministic apps it's an even bigger problem.

I knew a tech founder once who spent an hour lecturing us (his employees) on business ethics. He'd even written a little red book (like Mao) to codify his thoughts on how we should all behave.

Fast forward a few years and the guy flees the US after being charged with securities fraud. Spends the rest of his life living on his millions in a foreign country with no US extradition treaty.

Look, I love LLMs and even implement them for customers, but I am very sceptical about them 'replacing' ERP and CRP systems. What some AI folks don't seem to understand is that traditional ERP and CRP apps are completely driven by auditable business rules because they have to be. If you're running a company, there's no discretion at all about how money and other assets and liabilities are accounted for. It all has to be strictly according to the rules. This goes for most everything else - management are responsible for the business rules implemented in the system and they need to be precisely spelled out. Sure, AI can and should be used extensively for the human UI piece of it. To simplify getting data into and out of the system for example. But the engine inside and the database are all strictly rule governed and I definitely dont expect that to change anytime soon.

I'm surprised not to see anything about David Cope, who is the real master of AI produced music. Cope works mostly within the classical tradition, using AI to produce work in the style of great composers like Bach, Mozart etc. To my ear, it lacks the brilliance of the originals, but some of it is quite good. Certainly several orders of magnitude better than Suno's AI slop.

Gettier cases tell us something interesting about truth and knowledge. This is that a factual claim should depict the event that was the effective cause of the claim being made. Depiction is a picturing relationship: a correspondence between the words and a possible event (eg a cow in a field). Knowledge is when the depicted event was the effective cause of the belief. Since the paper mache cow was the cause of the belief, not a real cow, our intuitions tell us this is not normal knowledge. Therefore, true statements must have both a causal and depictional relationship with something in the world. Put another way, true statements implicitly describe a part of their own causal history.

I think you are asking the wrong question. Before you decide what to bring in house, you need to have an IT strategy That sets out what components of your desired solution should be built, and what should be bought. You need to have a view about which technologies you will use, and why. What is your end state IT architecture? Have you prioritised your objectives and requirements?

I understand your current systems are holding you back and there may be no appetite to get a bunch of expensive consultants in to answer all these questions. But just hiring a bunch of devs without a plan for what you want them to do is a recipe for disaster.

I know Orwell's work pretty well, and I read that sentence, and thought to myself: "Cant remember where he said anything like that, but what the hell, I haven't read everything Orwell ever wrote". So I just rolled with it.

The cognitive load to fact check everything is too great, so we decide which sources we think are reliable and just accept them. The solution is not to disbelieve everything you are told, but to accept that some of the facts you have not checked might be wrong, and be prepared to re-evaluate when contrary evidence appears.

Excellent. I agree totally. But trying to get adherence to standards in requirements documentation is like trying to nail jelly to the ceiling.

As others have said, rewrite the CV. It's not clear to me who you are and what kind of role would best suit you. Put a summary at the top making clear what you are best at and what kind of role you are looking for. Summarize your skills and achievments. Consider using simple graphics. eg List of tech skills rated with slide bar from expert to familiar understanding.

LLMs make writing code easier, but they don't make it unnecessary. "No-code" is mostly small consumer apps and POCs, but "low-code" is a big industry based on platform building. Much of the work of building and maintaining an enterprise system is (a)data model definition (b) UI/UX construction and (c) workflows. Low code platforms specialise in doing this in a structured, fast and maintainable way without code. LLMs are great but they aren't a magic bullet. Enterprise grade development requires a lot of structure and LLMs don't provide that. Low code provides a pre built, well documented structure, enabling mediocre developers like me to build reliable, professional grade enterprise apps.

Lots of different types of tools here. Some are no code, most are low code. There is no clear dividing line between them: they are all configurable platforms with the ability to incorporate code through APIs and add ons.

Some are just special purpose utilities (eg Postman). Some are niche products with a relatively limited set of use cases(eg Airtable), some are enterprise grade platforms which can be used to build huge systems (eg Salesforce, MS Power platform), and there is every type of gradation between these.

This is a multi billion dollar business, growing steadily. It's going to accelerate even faster as LLM Technologies are increasingly tuned to build code extensions.

The NFL theorem means nothing if all the learning tasks have a common underlying structure. In the real world, they do. The laws of physics and chemistry create emergent causal relationships. Any SSL learning algorithm that learns to exploit causal relationships will consistently perform well over a variety of real world tasks.

Don't know how accessible or affordable therapy is where you live, but you certainly need someone to talk to.

I had a midlife crisis in my 40's. My career was going nowhere, I was trapped in a job I hated. I felt I was wasting my life.

What kept me going was the sense of obligation I had to others who depend on me. If you have nothing or no one to whom you owe service, then you can change that. Find a partner, or volunteer yourself for charitable work. By helping others, you also help yourself.

I made a point of enjoying the little things in life. Exercise. Walk in the sun. Go fishing. Grow tomatoes. Buy a friend a drink. Whatever, just sneak the little pleasures when you can.

I set some goals. Not boring goals like to make money, goals to achieve something meaningful, that engaged my interests and (modest) abilities.

Things got better. My 50's and 60's were some of the most rewarding years of my life. (I'm 69 now). I achieved some things I'd never thought I could achieve, just by pluggin' away.

Life is hard. John Stuart Mill wrote that most people have a limited capacity for happiness. I think that is true. But the strange thing is that by relieving yourself of the obligation to feel happy, sometimes it happens anyway.

Good point. The essence of processing is the transformation and combination of signs. Processing a sign is a kind of energy transformation. The process takes input signs, performs some kind of energy transformation, and produces a new sign, with different signification, which may be a better or more useful signifier for some purpose. Think NAND gates, or neurons. Example: if it looks like a duck and quacks like. duck, then it's probably a duck. A neuron summing 'looks like a duck' signal and 'quacks like a duck' signals is a more reliable signifier for ducks than either of the inputs. As for signs, see C S Peirce or Ruth Millikan for a clearer explanation. TLDR: Signs are relations between objects and processes (Sign, Signifier and Interpretant)

I'm 69, about to turn 70. I'm in a senior technical consulting role which often requires me to go on the tools. I work for a big, successful software firm that can afford to pay well. Most of my peers are 20+ years younger than me and quite smart, so I have to work hard to keep up. But I manage. Next year I will quit my job and 'retire'. In fact I'll be using my financial security to start up a SaaS. business. I have a bunch of ideas I've been wanting to work on. Working keeps me focused and connected. My rule is: stay interested, stay connected.

The Robodebt scandal in Australia didn't spill over to violence in the streets, but it came close. Robodebt was a debt collection system introduced by the Federal government to recover supposed overpayments of social security. It was flawed at every level, including software design, and resulted in nationwide protests. see https://www.canberratimes.com.au/story/6037228/robo-debt-cen...

I studied philosophy to postgraduate level before having to go out and get a real job. Now in late middle age, I am nearly recovered to the point where I have to agree with the author. I now understand that most philosophy, of ANY age or period, including the present, is useless. But there are counterexamples. Hume is still worth reading, on almost any subject. Many of his ideas have not aged at all. Reason as the slave of the passions, is vs ought, compatibilism on free will all come to mind, as do his deep insights on the connections between ideas (Resemblance, Contiguity, cause and effect). These positions are still held by scientifically literate people, and some of them, like the idea of connections between ideas, I have have actually found useful as guidelines for research in AI. And that's without mentioning his influence on political philosophy, which was profound. Hume's notion of the social contract continues to shape political debate in ways not many people realise.

I'm a 69 year old who made a conscious decision 12 years back to leave sales and management and return to my technical roots. Now I'm in a solution architect role where I'm expected to constantly stay abreast of the latest technologies. The pay is good. I don't have to do much coding, but I do some anyway to keep my hand in.

My advice to your Dad is to aim for a job in consulting and analysis rather than coding. Get some certifications in a SaaS platform like Dynamics or Salesforce or whatever. That way he can leverage his business experience and people skills. I've found that people who combine experience with solid knowledge of a software platform get quite a lot of interesting job offers, even if they are older. If he later finds out he has an aptitude and interest in development then he can go down that path after he has mastered the functional stuff.

In no particular order: - The siege of Krishnapur by J G Farrell - War and Peace by Leo Tolstoy - The Radetsky March, by Joseph Roth - The Leopard, by Giuseppe Tomasi di Lampedusa - Cloud Atlas by David Mitchell - Blood Meridian by Cormac McCarthy - Wolf Hall, by Hilary Mantel - 1984, by George Orwell - Never Mind, by Edward St Aubyn

Worth noticing that prosecution of foreign companies under Australian law requires the consent of the Federal Attorney General. This consent has been given, meaning that the Australian Govt agrees he has a prima facie case. Forrest has deep pockets and I hope he succeeds. FB's 'we're only a platform' schtick is wearing pretty thin.

I knew a guy once with a milder version of what appears to be a similar syndrome. We worked together, and he would go out for lunch and wolf down a whole large pizza. Afterwards, he would get very hot and sweaty. He was not particularly tall and skinny as a rake. He wasn't an athlete, had a sedentary job, yet he was eating 2-3 times normal calorie intake, maybe 5-6K calories/day at least. In all other respects he seemed a normal guy, who to my knowledge did not eat toddlers. There was clearly some medical reason for what was going on, but since he seemed normal in other respects as far as I know he never went to the doctor about it. I'm not medical and have no idea what his syndrome was, but I wonder if it could have had something to do with his mitochondrial functions. See https://astralcodexten.substack.com/p/shilling-for-big-mitoc... I'm not suggesting Tarrare was taking a drug that hadn't been invented yet, but that possibly he naturally had a natural version of the mitochondrial permeability syndrome that DNP induces.

Some remarks from Captain Obvious here.

The primary use case for crypto is simple: (1) People buy crypto because it keeps going up in value (2) It keeps going up in value because people keep buying it. (3) For why people keep buying Crypto, see (1).

That's the logic of a Ponzi scheme. All previous Ponzi schemes failed because either (a) sooner or later the scheme runs out of new suckers, and/or (b) the authorities intervene and shut it down, and/or (c) The founders cash out and 90% of the money just disappears into thin air.

The truly fascinating thing about crypto is that neither (a) or (b) appears likely in the near future. It's a Ponzi scheme with fungible shares.

For (c), as with other Ponzi schemes, the founders make a lot of money, the later entrants lose a lot of money. Sooner or later, the founders of BTC (for example) will start to cash out. At the moment, they seem to be cashing out gradually, so as not to cause a crash in their remaining holdings. That is one reason BTC prices are so volatile.

At the same time as the BTC founders are trying to find the optimum point in their cash out equation, there is another use case for crypto which continues to drive demand: crime and money laundering. Crypto is a great way to move money between jurisdictions. For some criminals, it is well worth the risk of volatility.

Ponzi logic and money laundering logic will continue to drive the price of established cryptos upwards. I suspect it might be quite a while before the game stops. That makes cryptos like BTC an interesting short term speculation. But to paraphrase Gurf Morlix: stay in too long, you're gonna get cut, and you're gonna bleed.

Pathetic site

Note that these numbers are from Italy, which is known to have extremely low intergenerational income mobility. See https://voxeu.org/article/intergenerational-mobility-across-... One key characteristic of high mobility countries is the availability of high quality public education (primary, secondary and tertiary). High mobility countries (eg Denmark, Canada, Australia) tend to have a large, low cost tertiary education sector where admission is strictly determined by independent measures of academic achievement. The old school tie counts for little or nothing in these societies.

Microsoft's Power Platform is a low code framework which works well, and generates a massive amount of revenue, as does Salesforce and some others. I recently designed and implemented a complex 500 user child protection application with PP that has been live for a year now. It was highly successful, and the time and cost taken to deliver it was far less than the cost of a hand written solution. That said, there is still quite a lot of custom code required for most enterprise level solutions even with the most mature low code platforms. Low code is not a panacea, and the same issues of how to represent requirements and design arise in the low code world as in the high code world. Low code platforms will continue to mature and improve. Maybe AI will catch them one day, but I'd be surprised if that happens anytime soon.