What We Misunderstand About Pre-Industrial Craftsmanship by Vic Tesolin
-
Its here to stay as long as VCs are okay with burning their cash on it, seems like they have a lot to be okay with running on loss making AI products
https://www.a16z.news/p/we-raised-15b-why
https://www.cnbc.com/2026/01/07/anthropic-funding-term-sheet-valuation.html
They have a comfortable runway to change workplace habits, practices and hiring for the next five years.
@impactology capital evaporated in 2008. It will do so again
-
AI has found a few niches.
1. Stock market manipulation
"Markets are full of predatory trading algorithms. They’ve evolved to hunt in packs." William Gibson
2. Election meddling
3. Automated disinformation
4. Union busting
5. Malign influence campaigns
6. Jailbreaking IP
7. Surveillance capitalism
8. Price fixing & rent fixing
9. Monopolization@Npars01 @impactology Automated trading on various markets makes use of AI. I learned a bit about this when I had consulted for a major investment house on their AI tech. They told me 11% of all trades were automatic. It is very complicated to explain how it worked, but it was extremely alarming to me because there were no effective guardrails to prevent emergent escalation effects and market collapse. This was over 10 years ago so maybe it is safer, but I don’t think so.
-
@impactology capital evaporated in 2008. It will do so again
@urlyman @Npars01 By then the damage will be done, people can't stop their lives and wait for that. You have to live and adapt
Already recruiters and hiring mangers are clueless on evaluating skills, its going to get even worse when they insist on evaluating AI tool use. Ones who already are senior level don't have to worry because they already have a track record and prestige, ones that are entering the field now are scrambling
-
For businesses its really simple to judge its value : if it saves more money than it costs to use it, otherwise dump it
@impactology @Npars01 There are some domains where decision-making requires processing massive amounts of data and/or require decisions in fractions of a second. People cannot handle this work, they can’t even be in the loop. These are the problem domains most suited to AI-based automation. The value depends a lot on what the problem is that AI is solving. There is no question AI is delivering value well above its cost. That means it will be adopted more and more widely. No doubt in my mind.
-
@urlyman @Npars01 By then the damage will be done, people can't stop their lives and wait for that. You have to live and adapt
Already recruiters and hiring mangers are clueless on evaluating skills, its going to get even worse when they insist on evaluating AI tool use. Ones who already are senior level don't have to worry because they already have a track record and prestige, ones that are entering the field now are scrambling
@impactology I’m lucky in that I don’t because I’m in my 60s. My generation and the one after me is pumping a huge river of shit on young people. I hope they find ways to stay out of its strongest currents, because submitting to them will wash them away
-
@impactology @Npars01 There are some domains where decision-making requires processing massive amounts of data and/or require decisions in fractions of a second. People cannot handle this work, they can’t even be in the loop. These are the problem domains most suited to AI-based automation. The value depends a lot on what the problem is that AI is solving. There is no question AI is delivering value well above its cost. That means it will be adopted more and more widely. No doubt in my mind.
Manufacturing, logistics, aviation, food safety and finance right? For example Import-export trade compliance software for shipping?
https://mastodon.social/@impactology/115912577302179967
Or any of these domains below?
-
@urlyman @Npars01 By then the damage will be done, people can't stop their lives and wait for that. You have to live and adapt
Already recruiters and hiring mangers are clueless on evaluating skills, its going to get even worse when they insist on evaluating AI tool use. Ones who already are senior level don't have to worry because they already have a track record and prestige, ones that are entering the field now are scrambling
@impactology @urlyman @Npars01 being senior holds no weight if you aren’t toeing the line
-
@impactology I’m lucky in that I don’t because I’m in my 60s. My generation and the one after me is pumping a huge river of shit on young people. I hope they find ways to stay out of its strongest currents, because submitting to them will wash them away
Specifically your generation's C-suite folks and they don't even have to worry about pumping shit coz they'll get their kids prestigious internships via the side door anyway.
-
@impactology @urlyman @Npars01 being senior holds no weight if you aren’t toeing the line
@fasterandworse @urlyman @Npars01
Sure but you have a referral network to vouch for you for a job opening so you can bypass the getting screened from 1000 of applicants based on shiny college logos part and get straight to the interview
-
@fasterandworse @urlyman @Npars01
Sure but you have a referral network to vouch for you for a job opening so you can bypass the getting screened from 1000 of applicants based on shiny college logos part and get straight to the interview
@impactology @urlyman @Npars01 I hear you but I'm referring to the roles themselves. I admire your positive perspective but the word "adapt" is the problematic load-bearer here.
-
@impactology @urlyman @Npars01 I hear you but I'm referring to the roles themselves. I admire your positive perspective but the word "adapt" is the problematic load-bearer here.
@fasterandworse @urlyman @Npars01
Ah got it, yes you're right.
Stephen you guys should sell courses, seriously all that tacit expertise on framing problems, finding errors, preventing them and fixing it. All that pre-ai needs to be preserved and taught. The books aren't cutting it because they are too template-y
-
-
@fasterandworse @urlyman @Npars01
Ah got it, yes you're right.
Stephen you guys should sell courses, seriously all that tacit expertise on framing problems, finding errors, preventing them and fixing it. All that pre-ai needs to be preserved and taught. The books aren't cutting it because they are too template-y
@fasterandworse @urlyman @Npars01
Vitaly Friedman for example is totally killing it
https://x.com/vitalyf/status/2007729848891126098
We need you seniors to be loud and open on social media talking about solving complex problems, because currently its being dominated by AI hype bros and landing page designers
-
@fasterandworse @urlyman @Npars01
Vitaly Friedman for example is totally killing it
https://x.com/vitalyf/status/2007729848891126098
We need you seniors to be loud and open on social media talking about solving complex problems, because currently its being dominated by AI hype bros and landing page designers
@impactology the future of UI is less complex, but we will explore making it more complicated first
-
Manufacturing, logistics, aviation, food safety and finance right? For example Import-export trade compliance software for shipping?
https://mastodon.social/@impactology/115912577302179967
Or any of these domains below?
@impactology @Npars01 I recently learned about a similar problem in health care insurance billing. It has become so complicated and time consuming that (US) providers now use specialists who know what works, and that is not the same as what is correct. But you’d have to match improvements on the insurance side also because now they’ve accommodated to getting wrong answers. It’s a costly mess.
-
Manufacturing, logistics, aviation, food safety and finance right? For example Import-export trade compliance software for shipping?
https://mastodon.social/@impactology/115912577302179967
Or any of these domains below?
@impactology @Npars01 I’m familiar with use of #AI in the first three domains you mention. One of my own inventions was licensed by a big pharmaceutical company to automate a factory. In general, logistics and supply chain in nearly every industry are fertile areas as well. You don’t hear about those successful applications because the terminology for the tech used is typically more precise (e.g., “constraint satisfaction”) and thus harder to explain to non-experts. This is antithetical to hype.
-
@impactology @Npars01 I recently learned about a similar problem in health care insurance billing. It has become so complicated and time consuming that (US) providers now use specialists who know what works, and that is not the same as what is correct. But you’d have to match improvements on the insurance side also because now they’ve accommodated to getting wrong answers. It’s a costly mess.
Based on the numbers of use cases for AI, frauds & scams appear to be the most used.
More than one pundit has observed that Silicon Valley isn't in the "innovation" industry.
It's in the "fake it, until you make it" smoke & mirrors vaporware industry.
It sucks up venture capital cash and produces lies more than products.
https://www.businessinsider.com/venture-capital-big-tech-antitrust-predatory-pricing-uber-wework-bird-2023-7https://www.sacbee.com/opinion/op-ed/article312606310.html
https://www.panewslab.com/en/articles/fbcc6570-210b-49d4-946e-27b46c4a836d
-
Based on the numbers of use cases for AI, frauds & scams appear to be the most used.
More than one pundit has observed that Silicon Valley isn't in the "innovation" industry.
It's in the "fake it, until you make it" smoke & mirrors vaporware industry.
It sucks up venture capital cash and produces lies more than products.
https://www.businessinsider.com/venture-capital-big-tech-antitrust-predatory-pricing-uber-wework-bird-2023-7https://www.sacbee.com/opinion/op-ed/article312606310.html
https://www.panewslab.com/en/articles/fbcc6570-210b-49d4-946e-27b46c4a836d
@Npars01 @impactology Vaporware is a constant problem, from my POV. I’d call that business fraud. AI is no different than many other technologies or business models based on hype. Unfortunately, hype sometimes does pay off in the short term so it persists.
-
@impactology @Npars01 I’m familiar with use of #AI in the first three domains you mention. One of my own inventions was licensed by a big pharmaceutical company to automate a factory. In general, logistics and supply chain in nearly every industry are fertile areas as well. You don’t hear about those successful applications because the terminology for the tech used is typically more precise (e.g., “constraint satisfaction”) and thus harder to explain to non-experts. This is antithetical to hype.
https://gijn.org/stories/gijc25-uncover-transnational-scam-operations/
https://www.fastcompany.com/91460661/deepfakes-business-supply-chain-risk-executives
https://www.cfo.com/news/advanced-generative-ai-cyber-fraud-rose-118-last-year/739157/
https://www.reuters.com/practical-law-the-journal/transactional/ai-supply-chain-2025-06-01/
https://www.globaltrademag.com/supply-chain-instability-opens-the-door-to-ai-powered-fraud/
https://securitybrief.com.au/story/australian-businesses-lose-aud-2-03-billion-as-ai-scams-surge
https://www.paymentsjournal.com/the-payment-process-the-supply-chains-most-overlooked-cyber-risk/
-
@Npars01 @impactology Vaporware is a constant problem, from my POV. I’d call that business fraud. AI is no different than many other technologies or business models based on hype. Unfortunately, hype sometimes does pay off in the short term so it persists.
Knew which side AI initiatives had joined when Koch announced funding for forced user adoption for "economic mobility".
https://www.forbes.com/sites/mattdurot/2025/07/17/bill-gates-charles-koch-and-three-other-billionaires-are-giving-1-billion-to-enhance-economic-mobility-in-the-us/Charles Koch spent the last 50 years immiserating the middle & working classes, ending democracy, & frying the planet.
He & his ilk are completely indifferent to anyone but themselves.
If he's spending money on coercing AI adoption, we know why. More of the same, but faster.
https://www.nytimes.com/2025/12/20/us/politics/koch-network-2024-election-trump.html
