Alex

520 posts

Alex banner
Alex

Alex

@FullyAlexD

Building @ Fulfilled Wealth | Prev @Microsoft eng

Toronto, Ontario Katılım Kasım 2025
438 Takip Edilen201 Takipçiler
Alex
Alex@FullyAlexD·
@luke_metro Just because you frame a narrative like that won't actually make it true. Its just pretending that the first impacted wont be domestic workers when everyone knows it will be. Needs to be backed up with empirical proof that this will be expansive and increase the worker base.
English
0
0
6
568
Luke Metro
Luke Metro@luke_metro·
The narrative for AI x manufacturing really needs to be that almost all the American jobs have already been shipped overseas so we’re replacing Chinese workers with American robots+technicians Otherwise we get … this
Bernie Sanders@BernieSanders

Jeff Bezos, worth $234 billion, plans to replace 600,000 Amazon workers with robots. Now, he wants to spend $100 billion to fully automate not just his warehouses, but factories in the U.S & other countries. Oligarchs are waging all out war against workers. FIGHT BACK.

English
15
31
472
32.1K
Alex
Alex@FullyAlexD·
These numbers were available for projection when that report was written, thats why they say they used straight line estimations and its beyond 2026 is where they have less clarity. Even then, youre being generous. Dario said just last month that 4.6 is giving them roughly a 25% productivity boost internally. Half of AI2027s productivity boost formula was in compute scale alone.
English
0
0
0
269
Psyho
Psyho@FakePsyho·
Seems that AI 2027 (ridiculed for "impossible" timelines) severely underestimated the speed of progress in late 2025 / early 2026: - AI coding agents have a much greater impact than the projected speedups - OpenAI alone already matched the revenue estimate two months earlier ($25B in Feb); if we combine revenue from all frontier labs, we've probably already matched the Jan 2027 estimate ($55B) I wouldn't be that much surprised if the authors revert to their original timelines at some point
Psyho tweet media
English
28
33
451
28.8K
Pavel Snajdr
Pavel Snajdr@PavelSnajdr·
@tekbog > if the global economy and the supply chain gets disrupted that's a pretty big if
English
1
0
2
242
terminally onλine εngineer
terminally onλine εngineer@tekbog·
if the global economy and the supply chain gets disrupted i dont think we are getting the data centers up AGI officially delayed
English
17
8
186
12.7K
Alex
Alex@FullyAlexD·
@loganthorneloe Old study compared to SOTA but just to highlight the point: x.com/i/status/20337…
Amir Salihefendić@amix3k

In 2024, @ttunguz shared a great study on the use of AI in sales. The most striking part was the gap between perception and reality: people felt AI improved productivity by 25% to 75%, and some even claimed 500%, while the data showed changes ranging from -4% to 4%. There is something very off about how we perceive AI's impact on productivity compared to reality. I would love to see this study repeated in 2026, with a focus on software development as well.

English
0
0
0
24
Alex
Alex@FullyAlexD·
@loganthorneloe The issue is that the individual is a terrible judge of what actually constitutes productivity or velocity gain. Everyone thinks they've sped up against zero measurement baseline but if actually measured it often turns out to be pretty flat. Not aimed at you, just generalized
English
2
0
1
25
Logan Thorneloe
Logan Thorneloe@loganthorneloe·
I've found many good uses for LLMs with massive velocity gains. I've also found bad use cases that cause enormous velocity hits (that I wouldn't have expected!). It's becoming increasingly important to track the velocity gain/hit of an agent in its specific application.
David Cramer@zeeg

im fully convinced that LLMs are not an actual net productivity boost (today) they remove the barrier to get started, but they create increasingly complex software which does not appear to be maintainable so far, in my situations, they appear to slow down long term velocity

English
1
2
2
455
Alex
Alex@FullyAlexD·
I dont know what youre working on so its impossible for me to debate that. Id just say that based on what we've seen all over the industry, it remains to be seen that this is a universal experience even from highly skilled teams. Both anecdotal and from industry wide reliability & consumer trends. Doesnt stop me from using LLMs/agents religiously but trying to be realistic about where the bar currently sits.
English
0
0
0
136
Ryan Brewer
Ryan Brewer@ryanbrewer·
Again I’d point to the tweet I linked. There were always startups that wrote bad code, and they usually failed. It’s the same now! It’s on you to be able to use these tools to accelerate, and usually requires skilled engineering to receive the benefits (not writing unmaintainable slop). For my team and most teams @OpenAI I’d say this has been a massive acceleration in productivity
English
1
0
0
152
Alex
Alex@FullyAlexD·
@ryanbrewer Tail where economic benefit is metered** Ie. Asking a dev if they are more productive now is an irrelevant exercise. Theyre almost never in a position to measure their own productivity.
English
0
0
0
22
Alex
Alex@FullyAlexD·
Net-productive as a founder, yes. I built my fintech startup mostly manually early 2025. Was able to replicate the design patterns to build quicker. Startup was greenfield. Under 10k MAU. However, founders ive met on the otherside of the bottleneck are now experiencing the pains from going hands off without guardrails too soon. Re-writes, significant perf issues, churn from loss of trust. Productivity isnt measured by just output alone. You need to measure the tail where the economic benefit is actually measured.
English
2
0
0
164
Alex
Alex@FullyAlexD·
@ryanbrewer "Net positive" is doing the heavy lifting in his statement. He even addresses your quoted response in his follow up comments.
English
1
0
4
152
The Holy Ghost
The Holy Ghost@Theolojism·
@theCTO Why don’t you tell him yourself, oh wait because he doesn’t care
English
1
0
2
2.5K
Alex
Alex@FullyAlexD·
@0xen1 @thattallguy It polls at under 20% in a wide set of "normie" polls. Nothing polls at sub 20% in the US.
English
1
0
10
80
Oxen
Oxen@0xen1·
@thattallguy i don't think the general public hates it all that much. the creative/designer/urbanite general public does, but that's a bit of a bubble.
English
4
0
62
3.1K
Jess 🌱
Jess 🌱@thattallguy·
I think folks on this app are greatly underestimating how much the general public hates AI. Like, haaaaaaates it.
English
622
430
7.1K
187.3K
Alex
Alex@FullyAlexD·
@miratechgeek @ben_j_todd Industrial revolution clocked more like 100-200% over a longer period of time. Ive never heard anyone say something like 1% from the literal creation of productivity machines. Ask any LLM to benchmark the general economy in Britian through it.
English
0
0
0
25
FemMiraMancer
FemMiraMancer@miratechgeek·
@ben_j_todd 10% productivity isn't 'normal.' Industrial Revolution clocked 1-2%. But fair. If AGI, where's the science acceleration?
English
1
0
2
202
Benjamin Todd
Benjamin Todd@ben_j_todd·
If AI progress stopped now, it would be a normal technology. One-off 5-10% productivity growth. Some routine white collar tasks automated. We chat to AI tools a lot. But no big economic or scientific acceleration. Ergo we don't have AGI.
English
55
6
181
22.8K
Alex
Alex@FullyAlexD·
@kiaran_ritchie @MichaelNStruwig Have to agree here. Toughest problem to make agents both reliable + profitable (in personal finance niche) was that cheaper/OS models really lost it on tool calling. Trade off between reliability and affordability.
English
0
0
0
63
Kiaran Ritchie
Kiaran Ritchie@kiaran_ritchie·
I don't see how Anthropic, OpenAI or any of the model providers have any hope of defending their moats. And consequently, I think they're going to get wiped out. Right now, in early 2026 they have a meaningful advantage in terms of model capability. But far cheaper and open source models are not far behind. How long can they maintain a meaningful advantage? For the vast majority of use cases, we don't actually need much higher intelligence. It doesn't take 140 IQ to automate Turbotax or powerpoint. Eventually we will be saturated in cheap, local models that are "good enough". Of course some scientific labs and frontier research will always want the latest and greatest. But that market is orders of magnitude smaller than these company valuations can justify. What am I missing?
English
556
54
1.3K
251.7K
Alex
Alex@FullyAlexD·
@luke_metro @bryancsk Wasnt the average non founding tenure like 6 months? A lot of people didn't hit the cliff
English
0
0
4
149
Luke Metro
Luke Metro@luke_metro·
@bryancsk Their stock got pumped so much that they’re doing great regardless
English
3
0
185
25.7K
Bryan Cheong
Bryan Cheong@bryancsk·
I feel really bad for the xAI folks who worked six-seven days a week and gained 20 lbs each only in the end to leave it like this.
English
61
27
2K
354.5K
Thomas Eastham
Thomas Eastham@easthambuilds·
@dansemperepico it depends on where I am running it. if on my personal local computer then I babysit it, reviewing each permission request carefully. whenever possible I will run claude code and similar tools on a VM or dedicated untrusted machine... and let'r rip!
English
1
0
2
1.9K
Daniel Sempere Pico
Daniel Sempere Pico@dansemperepico·
You guys all run Claude Code with claude --dangerously-skip-permissions right? Because otherwise how in the world can you sit there accepting every single permission when building something?
English
475
22
2.2K
285.6K
Alex
Alex@FullyAlexD·
@dansemperepico I did until it deleted my C drive on two seperate occasions. Much safer to white list and build guardrails with hooks.
English
0
0
0
189
Alex
Alex@FullyAlexD·
@Principal_ADE @Steve8708 Even if its fractional, this review captures none of that fraction. There is zero capture/persistence or ownership beyond a static review.
English
0
0
0
382
Steve (Builder.io)
Steve (Builder.io)@Steve8708·
anthropic literally charged us $25 for this
Steve (Builder.io) tweet media
English
119
41
1.4K
264K
Alex
Alex@FullyAlexD·
Not seeing how this is a valid response? Staff engineer has deep product ownership and is where the org and users go when things go badly wrong. The tradeoff is that the staff engineer wont have a mental trail of what might have recently broken, so triage takes longer resulting in second order effects & losses?
English
1
0
6
2.5K
Fernando
Fernando@Principal_ADE·
@Steve8708 That’s how much a staff engineer would cost to review your pr and would write the same thing
English
6
0
91
21.9K
Alex
Alex@FullyAlexD·
@Rengle820 @sterlingcrispin The odds that these 1000+ line PRs still don't need a review by a senior engineer at this point are pretty slim so its additive costs not replacement. At least with smaller PRs you can say its one or the other.
English
1
0
4
773

Keşfet

@luke_metro @PavelSnajdr @tekbog @loganthorneloe @OpenAI @ryanbrewer @Theolojism @theCTO