Over the next 6 months, AI-driven cyber tools will detect a wide range of previously dormant security bugs across many systems, marking a temporary catch-up period as AI capability outpaces existing patches.
“I do think that over say the next 6 months we're going to have this call it one-time period of catching up where AI-driven cyber is going to be able to detect a whole range of of bugs that maybe have been dormant over the past 20 years across a wide range of systems.”
View on YouTube
Explanation
AI-driven vulnerability discovery tools did expand significantly through late 2025 and into 2026, broadly consistent with the trend described, though a clean-cut 'one-time catch-up period' bounded to exactly six months is difficult to isolate.
David Sacks
Partly Right
attribution: medium
00:12:54
aitech
OpenAI will release a similar cyber-vulnerability-detection capability within the next few weeks, following Anthropic's Mythos playbook.
“We know that like let's say the Chinese open source models like Kimmy K2, it's about 6 months behind. So we have a window here of maybe 6 months where we're still in this pre-release period... and uh I guess open AI is going to release a similar thing in the next few weeks.”
View on YouTube
Explanation
OpenAI did continue releasing frontier models with security-relevant capabilities in the following weeks/months, though a directly comparable named 'cyber vulnerability detection' product specifically mirroring Anthropic's approach was not clearly confirmed.
OpenAI's Spud model, its first Blackwell-trained model, will be released very soon ("any day now").
“Mythos and Spud, which is going to be out from OpenAI any day now, which is the first Blackwell trained model at OpenAI.”
View on YouTube
Explanation
No independently verifiable public confirmation of an OpenAI model specifically named 'Spud' being released was found.
OpenAI and Google will both adopt the same pre-release sandboxing/hardening approach that Anthropic used for Mythos before releasing their own frontier models.
“I think open AAI will end up doing the same thing. I think Google will end up doing the same thing.”
View on YouTube
Explanation
Frontier labs including OpenAI and Google did increasingly adopt more cautious pre-release safety testing and staged rollout practices through 2025-2026, generally consistent with the trend though not confirmed as directly copying Anthropic's specific approach.
Open-source language models will eventually win out, capturing roughly 90% of token usage and undercutting the entire frontier/closed-model space.
“these I believe open source is going to win the day on the large language models and take 90% of the token usage and I think the entire frontier model space could be undercut by open source”
View on YouTube
Explanation
Open-source models gained significant market share by 2025-2026 but did not capture roughly 90% of token usage; closed frontier models (OpenAI, Anthropic, Google) continued to command the majority of paid enterprise and consumer token volume.
Brad Gerstner
Partly Right
attribution: medium
00:38:36
aitechventure
At some point, a genuinely successful open-source pre-training project (for training frontier-scale models) will emerge.
“I tend to agree with you Jason that there is going to be at some point a very successful open source project for pre-training. Absolutely.”
View on YouTube
Explanation
Open-source pretraining projects (DeepSeek, Llama, Qwen) achieved significant success and competitiveness by 2025, partially matching the prediction, though whether any has become the clearly dominant success story remains debated.
David Sacks
Partly Right
attribution: medium
00:39:51
aitech
The share of code that is AI-generated (versus human-written) will rise from roughly 5% today to about 95%, likely within the next few years.
“it might be 5% today meaning 5% of the codes AI generated versus human generated I think it's going to 95% I mean I bet any amount of money on that the only question is when probably over the next few years”
View on YouTube
Explanation
AI-generated code share rose substantially through 2025-2026 with widespread adoption of coding agents, trending toward the predicted trajectory, though a confirmed 95% figure industry-wide was not independently verified.
David Sacks
Wrong
attribution: medium
00:40:18
aitechmarkets
The market for AI coding tokens could become monopolized by the current early leader, driven by data and training-token flywheel effects.
“So, I do believe that the market for coding tokens could be monopolized.”
View on YouTube
Explanation
The AI coding token market remained competitive among multiple providers (Anthropic/Claude, OpenAI, Google, and various coding-agent startups) rather than becoming monopolized by a single early leader.
Anthropic will exit the year at an $80-100 billion revenue run rate, while OpenAI also sees a revenue inflection driven by an imminent new model release.
“So I think when you look out toward the end of the year, I would not be shocked if you see Anthropic exiting this year at 80 to 100 billion in revenue. And by the way, doing it at the same time that OpenAI, who is also on the wave, they'll be releasing an incredible model in the next imminently. They're going to be on that wave and you're going to see an inflection in their revenues as well.”
View on YouTube
Explanation
Anthropic's revenue run rate grew rapidly through 2025 but ended the year well below the $80-100B range predicted (more in the range of several billion to low tens of billions), while OpenAI did see a substantial revenue inflection.
By the end of June, SpaceX will join Meta and Google in having a 'fortress balance sheet,' giving well-capitalized incumbents leverage over compute-constrained AI labs.
“I think by the end of June, SpaceX will also have a fortress balance sheet.”
View on YouTube
Explanation
No specific, independently verifiable confirmation of SpaceX reaching a formally described 'fortress balance sheet' status by the specific predicted date was found.
Brad Gerstner
Partly Right
attribution: medium
01:14:35
politicsconflict
A Ukraine-Russia peace deal is close, Venezuela negotiations are progressing well, and there will be significant news on Cuba.
“By the way, while everybody's focused on Iran, stay tuned. I think we're getting close to a deal on Ukraine, Russia, right? Venezuela is, you know, kind of going seemingly very well. I think there's also going to be news on Cuba.”
View on YouTube
Explanation
No comprehensive Russia-Ukraine peace deal was finalized shortly after this prediction, though negotiations did continue through 2025-2026; Venezuela and Cuba situations also saw incremental but not fully resolved developments.
Brad Gerstner
Partly Right
attribution: medium
01:14:58
marketspolitics
If the Iran ceasefire, Lebanon talks, Ukraine-Russia, and Venezuela situations are resolved by July 4th (America's 250th anniversary), the stock market could rally sharply.
“If you land the plane on those things heading into America 250 July 4th, the market could really take off.”
View on YouTube
Explanation
None of the Iran, Lebanon, Ukraine-Russia, or Venezuela situations were fully and simultaneously resolved by July 4, 2025, though the stock market did perform strongly overall through mid-to-late 2025.
The US and Israel will find an offramp from the current Iran conflict/tension, given the economic and geopolitical incentives on both sides.
“So all of that kind of tells me that we will find an offramp.”
View on YouTube
Explanation
The US, Israel, and Iran did eventually reach a ceasefire arrangement following the June 2025 conflict, consistent with finding an offramp.
Many tech companies have hit peak employment for their current business lines; while new jobs will be created elsewhere, existing roles will continue to shrink as companies do more with less headcount.
“I think for a lot of these tech companies, we've hit peak employment. We're going to create a tremendous number of new jobs, but for the existing jobs, these companies are all realizing they can do more with less.”
View on YouTube
Explanation
Many large tech companies did continue reducing or holding flat headcount for existing business lines through 2025-2026 while emphasizing AI-driven efficiency, broadly consistent with the prediction, though new AI-driven job creation elsewhere was less pronounced than implied.