BEGIN:VCALENDAR VERSION:2.0 PRODID:-//fru.dev//Predictions//EN CALSCALE:GREGORIAN METHOD:PUBLISH X-WR-CALNAME:Predictions: coming due X-WR-CALDESC:Target dates of open predictions about data\, AI and tech\, from predictions.fru.dev X-PUBLISHED-TTL:PT12H BEGIN:VEVENT UID:openai-intern-level-research-assistant-september-2026@predictions.fru. dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20260930 DTEND;VALUE=DATE:20261001 SUMMARY:Due: Sam Altman: OpenAI will have an intern-level AI research assi stant by September 2026. DESCRIPTION:OpenAI will have an intern-level AI research assistant by Sept ember 2026.\nTarget as stated: by September 2026.\nSource: https://techcr unch.com/2025/10/28/sam-altman-says-openai-will-have-a-legitimate-ai-rese archer-by-2028/\nhttps://predictions.fru.dev/predictions/openai-intern-le vel-research-assistant-september-2026 URL:https://predictions.fru.dev/predictions/openai-intern-level-research-a ssistant-september-2026 TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:zuckerberg-most-llama-code-written-by-ai@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20261029 DTEND;VALUE=DATE:20261030 SUMMARY:Due: Mark Zuckerberg: Within 12 to 18 months\, most of the code fo r Meta's Llama efforts will be written by AI. DESCRIPTION:Within 12 to 18 months\, most of the code for Meta's Llama eff orts will be written by AI.\nTarget as stated: sometime in the next 12 to 18 months.\nSource: https://www.dwarkesh.com/p/mark-zuckerberg-2\nhttps: //predictions.fru.dev/predictions/zuckerberg-most-llama-code-written-by-a i URL:https://predictions.fru.dev/predictions/zuckerberg-most-llama-code-wri tten-by-ai TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:amodei-powerful-ai-as-early-as-2026@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20261231 DTEND;VALUE=DATE:20270101 SUMMARY:Due: Dario Amodei: Powerful AI could arrive as early as 2026. DESCRIPTION:Powerful AI could arrive as early as 2026.\nTarget as stated: as early as 2026.\nSource: https://darioamodei.com/machines-of-loving-gra ce\nhttps://predictions.fru.dev/predictions/amodei-powerful-ai-as-early-a s-2026 URL:https://predictions.fru.dev/predictions/amodei-powerful-ai-as-early-as -2026 TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:altman-systems-that-find-novel-insights-2026@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20261231 DTEND;VALUE=DATE:20270101 SUMMARY:Due: Sam Altman: 2026 will likely see AI systems that can figure o ut novel insights. DESCRIPTION:2026 will likely see AI systems that can figure out novel insi ghts.\nTarget as stated: 2026.\nSource: https://blog.samaltman.com/the-ge ntle-singularity\nhttps://predictions.fru.dev/predictions/altman-systems- that-find-novel-insights-2026 URL:https://predictions.fru.dev/predictions/altman-systems-that-find-novel -insights-2026 TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:anthropic-powerful-ai-late-2026-early-2027@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20270331 DTEND;VALUE=DATE:20270401 SUMMARY:Due: Anthropic: Powerful AI systems will emerge in late 2026 or ea rly 2027. DESCRIPTION:Powerful AI systems will emerge in late 2026 or early 2027.\nT arget as stated: late 2026 or early 2027.\nSource: https://www.anthropic. com/news/anthropic-s-recommendations-ostp-u-s-ai-action-plan\nhttps://pre dictions.fru.dev/predictions/anthropic-powerful-ai-late-2026-early-2027 URL:https://predictions.fru.dev/predictions/anthropic-powerful-ai-late-202 6-early-2027 TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:altman-real-world-robots-2027@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20271231 DTEND;VALUE=DATE:20280101 SUMMARY:Due: Sam Altman: 2027 may see robots that can do tasks in the real world. DESCRIPTION:2027 may see robots that can do tasks in the real world.\nTarg et as stated: 2027.\nSource: https://blog.samaltman.com/the-gentle-singul arity\nhttps://predictions.fru.dev/predictions/altman-real-world-robots-2 027 URL:https://predictions.fru.dev/predictions/altman-real-world-robots-2027 TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:openai-automated-ai-researcher-2028@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20281231 DTEND;VALUE=DATE:20290101 SUMMARY:Due: Sam Altman: OpenAI will have a fully automated AI researcher by 2028. DESCRIPTION:OpenAI will have a fully automated AI researcher by 2028.\nTar get as stated: by 2028.\nSource: https://techcrunch.com/2025/10/28/sam-al tman-says-openai-will-have-a-legitimate-ai-researcher-by-2028/\nhttps://p redictions.fru.dev/predictions/openai-automated-ai-researcher-2028 URL:https://predictions.fru.dev/predictions/openai-automated-ai-researcher -2028 TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:legg-50-percent-chance-of-agi-by-2028@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20281231 DTEND;VALUE=DATE:20290101 SUMMARY:Due: Shane Legg: There is a 50% chance of AGI by 2028. DESCRIPTION:There is a 50% chance of AGI by 2028.\nTarget as stated: by 20 28 (50% chance).\nSource: https://www.dwarkesh.com/p/shane-legg\nhttps:// predictions.fru.dev/predictions/legg-50-percent-chance-of-agi-by-2028 URL:https://predictions.fru.dev/predictions/legg-50-percent-chance-of-agi- by-2028 TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:huang-agi-by-tests-within-5-years@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20290319 DTEND;VALUE=DATE:20290320 SUMMARY:Due: Jensen Huang: AGI\, defined as passing a set of tests better than most people\, will arrive within 5 years. DESCRIPTION:AGI\, defined as passing a set of tests better than most peopl e\, will arrive within 5 years.\nTarget as stated: within 5 years.\nSourc e: https://techcrunch.com/2024/03/19/agi-and-hallucinations/\nhttps://pre dictions.fru.dev/predictions/huang-agi-by-tests-within-5-years URL:https://predictions.fru.dev/predictions/huang-agi-by-tests-within-5-ye ars TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:kurzweil-human-level-intelligence-by-2029@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20291231 DTEND;VALUE=DATE:20300101 SUMMARY:Due: Ray Kurzweil: Computers will reach human-level intelligence b y 2029. DESCRIPTION:Computers will reach human-level intelligence by 2029.\nTarget as stated: by 2029.\nSource: https://www.theguardian.com/technology/arti cle/2024/jun/29/ray-kurzweil-google-ai-the-singularity-is-nearer\nhttps:/ /predictions.fru.dev/predictions/kurzweil-human-level-intelligence-by-202 9 URL:https://predictions.fru.dev/predictions/kurzweil-human-level-intellige nce-by-2029 TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:amodei-half-of-entry-level-white-collar-jobs@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20300528 DTEND;VALUE=DATE:20300529 SUMMARY:Due: Dario Amodei: AI could eliminate half of entry-level white-co llar jobs and push unemployment to as much as 20% within one to five year s. DESCRIPTION:AI could eliminate half of entry-level white-collar jobs and p ush unemployment to as much as 20% within one to five years.\nTarget as s tated: in the next one to five years.\nSource: https://www.cnn.com/2025/0 5/29/tech/ai-anthropic-ceo-dario-amodei-unemployment\nhttps://predictions .fru.dev/predictions/amodei-half-of-entry-level-white-collar-jobs URL:https://predictions.fru.dev/predictions/amodei-half-of-entry-level-whi te-collar-jobs TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:altman-superintelligence-in-a-few-thousand-days@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20321210 DTEND;VALUE=DATE:20321211 SUMMARY:Due: Sam Altman: Superintelligence could arrive in a few thousand days. DESCRIPTION:Superintelligence could arrive in a few thousand days.\nTarget as stated: in a few thousand days (about 3\,000 days\, read as late 2032 ).\nSource: https://ia.samaltman.com/\nhttps://predictions.fru.dev/predic tions/altman-superintelligence-in-a-few-thousand-days URL:https://predictions.fru.dev/predictions/altman-superintelligence-in-a- few-thousand-days TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:ai-2027-superhuman-ai-impact-over-next-decade@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20350403 DTEND;VALUE=DATE:20350404 SUMMARY:Due: Daniel Kokotajlo: The impact of superhuman AI over the next d ecade will exceed that of the Industrial Revolution. DESCRIPTION:The impact of superhuman AI over the next decade will exceed t hat of the Industrial Revolution.\nTarget as stated: over the next decade .\nSource: https://ai-2027.com/\nhttps://predictions.fru.dev/predictions/ ai-2027-superhuman-ai-impact-over-next-decade URL:https://predictions.fru.dev/predictions/ai-2027-superhuman-ai-impact-o ver-next-decade TRANSP:TRANSPARENT END:VEVENT BEGIN:VEVENT UID:karpathy-agents-take-a-decade@predictions.fru.dev DTSTAMP:20260925T003838Z DTSTART;VALUE=DATE:20351017 DTEND;VALUE=DATE:20351018 SUMMARY:Due: Andrej Karpathy: Making AI agents work like an employee will take about a decade. DESCRIPTION:Making AI agents work like an employee will take about a decad e.\nTarget as stated: about a decade.\nSource: https://www.dwarkesh.com/p /andrej-karpathy\nhttps://predictions.fru.dev/predictions/karpathy-agents -take-a-decade URL:https://predictions.fru.dev/predictions/karpathy-agents-take-a-decade TRANSP:TRANSPARENT END:VEVENT END:VCALENDAR