I can accept the author's premise that LLMs double coding productivity today.
Semiconductors following Moore's Law increased their density, which is a reasonable metric for effectiveness, by 40% a year for several decades. Leaving aside projections of AGI in two years, it's still a substantial impact if we can figure out how to increase coding productivity by 40% a year for two or three decades. I don't think that gets us to AGI or the Singularity, but it's similar to the impact of the steam engine, steel, or electricity.
These are "normal technologies" that were transformational, and that may be the path we are on with LLMs.
tl;dr: Google DeepMind's AlphaEvolve improved the speed of Substrate's computational lithography stack by 680% and reduced compute costs by 97%, enabling the simulation and printing of the most demanding layers in advanced semiconductor devices.
Key take-away: "Domain knowledge makes you better at using LLMs. [...] Human expertise will continue to be useful even as models get stronger. The human is the bottleneck for many tasks,not the model: the difficult part is communicating to the model exactly what kind of solution the human wants. The information is “in the model” already, but it takes a very smart human to pull it out."
tldr; Here are five signs your startup may still be a hobby:
1. You are not making real offers that include a price, delivery date, and a promised result.
2. You rely on a spouse, significant other, inheritance, savings, or a tolerable day job for support, but you have not shared a credible plan for reaching break-even and then profitability.
3. You have not defined working hours, milestones, and what you expect to accomplish in a specific timeframe.
4. You are only getting feedback from people you already know.
5. You have not set up a separate bank account, business structure, and expense log.
"While reviewing a list titled “Retractions by Nobel Prize winners” (https://retractionwatch.com/retractions-by-nobel-prize-winne...) compiled by the industry watchdog Retraction Watch, historian Yves Gingras found that two essays by physics pioneer Max Planck had been retracted. Gingras, a historian at the Université du Québec à Montréal, found that the papers were retracted decades after their publication in 1940 and 1942 in Die Naturwissenschaften. [...]
Tim Kersjes, the publisher’s head of research integrity, explained that the “decision to retract the papers was a human error and we can confirm that no software or ‘bot’ would have been involved in the process. Unfortunately, our records are limited, as the retractions occurred in 2011, the individuals involved have mostly left the company and the systems by which we record integrity matters have since changed.”"
I don't care if it's trivial to implement and impossible to bypass: it's an effort to eliminate anonymous Internet browsing/commenting because everyone over 16 has to submit ID as well. Its the end of free speech on the part of the Internet the UK controls.
Context: the government has objectively become increasingly authoritarian, with the partial elimination of jury trials, the criminalisation of peaceful protest, the use of anti-terror sentencing laws for activities that are clearly not terrorist, and other actions which set up ideal conditions for an oppressive dictatorship.
It's hard to take the idea that this is about concern for teens seriously when the PM bypassed civil service vetting norms to make a known friend of Epstein ambassador to the US.
I'll believe this is actually about protecting children when they do anything to address the myriad of other issues young people today are facing. So far that doesn't seem to be happening.
Funnily, I'm also not seeing any talk about holding the social media companies themselves accountable for any of the damage they've done to society.
That was my experience as well. Paid for it for a year, it was a clean presentation of a subset of Google results, but when I was really looking for something I would fall back on Google. It's a shame their execution does not match their brand promise.
I don't even think it's an issue with their execution. Google just has a moat -- a huge index they won't give to anyone else, developed over nearly three decades, including multiple Internet eras already past -- which would take an inordinate amount of investment (including multiple more decades) to even hope to recreate.
Kagi probably won't get there -- in fact it's likely no one else even can get there -- but they're already somewhere and should only continue to get better.
Semiconductors following Moore's Law increased their density, which is a reasonable metric for effectiveness, by 40% a year for several decades. Leaving aside projections of AGI in two years, it's still a substantial impact if we can figure out how to increase coding productivity by 40% a year for two or three decades. I don't think that gets us to AGI or the Singularity, but it's similar to the impact of the steam engine, steel, or electricity.
These are "normal technologies" that were transformational, and that may be the path we are on with LLMs.
reply